# OpenCode

Published articles for OpenCode.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Who Is Behind the Harness? Verifying Model Identity in Coding Agents through Behavioral Fingerprinting

DevFeed: [Who Is Behind the Harness? Verifying Model Identity in Coding Agents through Behavioral Fingerprinting](<https://devfeed.tech/articles/who-is-behind-the-harness-verifying-model-identity-in-coding-agents-through-behavioral-fingerprinting-142116.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2609.28559>)

Author: Chuyi Wang, Xiaohui Xie, Tongze Wang, Fangchen Luo, Yong Cui

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [LLM security](<https://devfeed.tech/topics/llm-security.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [AI-assisted coding](<https://devfeed.tech/topics/ai-assisted-coding.md>), [test](<https://devfeed.tech/topics/test.md>)

Tags: [actions](<https://devfeed.tech/tags/actions.md>), [agents](<https://devfeed.tech/tags/agents.md>), [black-box](<https://devfeed.tech/tags/black-box.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [execution](<https://devfeed.tech/tags/execution.md>), [failure-recovery](<https://devfeed.tech/tags/failure-recovery.md>), [harness](<https://devfeed.tech/tags/harness.md>), [identity](<https://devfeed.tech/tags/identity.md>), [lidar](<https://devfeed.tech/tags/lidar.md>), [llm](<https://devfeed.tech/tags/llm.md>), [llms](<https://devfeed.tech/tags/llms.md>), [model](<https://devfeed.tech/tags/model.md>), [model-weights](<https://devfeed.tech/tags/model-weights.md>), [models](<https://devfeed.tech/tags/models.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [openhands](<https://devfeed.tech/tags/openhands.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [security](<https://devfeed.tech/tags/security.md>), [top](<https://devfeed.tech/tags/top.md>), [verification](<https://devfeed.tech/tags/verification.md>)

### AI overview

LIDAR identifies coding-agent models by actively probing their runtime decisions rather than relying on final text or token distributions. Three probe pairs test post-edit verification, recovery from transient failures, and resolution of specification-test conflicts. It represents execution trajectories with instance-level and distribution-level features, then compares them with clean references using a lightweight probabilistic identifier. The method needs no access to weights, logits, or provider internals, and uses separate references for OpenCode, OpenHands, and mini-swe-agent. Across 24 models from eight families, closed-set Top-1 accuracy ranged from 91.39% to 98.89%, exceeding three baselines in each harness. Including unregistered models from the same families reduced balanced accuracy to 79.58%-80.42%.

### Source excerpt

arXiv:2609.28559v2 Announce Type: replace-cross Abstract: LLMs increasingly operate through coding-agent harnesses that inspect repositories, invoke tools, and modify files. Substituting the model behind such an agent can therefore change security-relevant decisions, including whether it verifies changes or recovers safely from failures. Existing LLM fingerprints largely infer identity from direct text or token distributions. In coding agents, these signals are mediated by system instructions, controller logic, tools, and execution feedback, limiting their transfer. We present LIDAR (LLM Identification from Decisions and Actions at Runtime), an active black-box fingerprinting method for coding-agent execution. Three coding probe pairs expose post-edit verification, transient-failure recovery, and specification--test conflict resolution under controlled changes. LIDAR represents the resulting trajectories with complementary instance-level and distribution-level features and compares them with clean references using a lightweight probabilistic identifier. It requires no access to model weights, logits, or provider internals. We evaluate 24 models from eight families on OpenCode, OpenHands, and mini-swe-agent, fitting separate references for each harness. Closed-set Top-1 accuracy ranges from 91.39\% to 98.89\% across the three harnesses, exceeding all three baselines in each setting. When the audit also includes unregistered models from the same families, balanced accuracy ranges from 79.58\% to 80.42\%. These results show that agent execution behavior provides model-identity evidence beyond final outputs.

## What Does a Harness Buy? Tokens, Mostly

DevFeed: [What Does a Harness Buy? Tokens, Mostly](<https://devfeed.tech/articles/what-does-a-harness-buy-tokens-mostly-142118.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.04433>)

Author: Yangze Liu, Zhongyi Han

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [AI Context Management](<https://devfeed.tech/topics/ai-context-management.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [code](<https://devfeed.tech/tags/code.md>), [coding](<https://devfeed.tech/tags/coding.md>), [context](<https://devfeed.tech/tags/context.md>), [context-management](<https://devfeed.tech/tags/context-management.md>), [cost](<https://devfeed.tech/tags/cost.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [harness](<https://devfeed.tech/tags/harness.md>), [language-model](<https://devfeed.tech/tags/language-model.md>), [leaderboards](<https://devfeed.tech/tags/leaderboards.md>), [model](<https://devfeed.tech/tags/model.md>), [models](<https://devfeed.tech/tags/models.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [pass](<https://devfeed.tech/tags/pass.md>), [points](<https://devfeed.tech/tags/points.md>), [prompt](<https://devfeed.tech/tags/prompt.md>), [swe-bench](<https://devfeed.tech/tags/swe-bench.md>), [system](<https://devfeed.tech/tags/system.md>), [tokens](<https://devfeed.tech/tags/tokens.md>), [tools](<https://devfeed.tech/tags/tools.md>)

### AI overview

Across seven models tested with Claude Code, mini-SWE-agent, and OpenCode, no harness leads consistently. On SWE-bench Verified and two other benchmarks, bounded differences between harnesses are 2-5 points, and swapping harnesses changes about as many task outcomes as rerunning the same setup. Some model-harness pairings show larger, model-dependent effects: OpenCode trails by 8-9 points on one model and leads by 14 on another. Cost differences also vary by model, ranging from under 10% to 3-4 times. The authors conclude that harness comparisons should be reported alongside reruns; the abstract does not establish a universally better harness.

### Source excerpt

arXiv:2610.04433v2 Announce Type: replace-cross Abstract: A coding agent is a language model inside a harness: the system prompt, the tools, and the context management that let a chat model work in a repository. Most of the engineering on these agents goes into the harness, and vendors and leaderboards credit it with score gains. Earlier studies have fixed the model and swapped the harness, but have not measured whether the swap changes the pass rate more than running the same harness again. We run seven models through three production harnesses, Claude Code, mini-SWE-agent, and OpenCode, on SWE-bench Verified and two benchmarks of other kinds, and rerun them unchanged. No harness leads consistently across models. Where the data can bound the difference between two harnesses, the bound is 2 to 5 points, and a harness swap changes about as many task outcomes as a rerun. The exceptions are harness effects in particular pairings of model and harness, whose direction changes with the model: OpenCode trails by 8 to 9 points on one model and leads by 14 on another. Cost differences between harnesses also depend on the model, from under 10% to 3 to 4 times. Which harness is better therefore has no answer apart from the model, and a harness difference should be reported next to a rerun.

## YUL: Bringing Software Supply Chain Security to Coding Agents

DevFeed: [YUL: Bringing Software Supply Chain Security to Coding Agents](<https://devfeed.tech/articles/yul-bringing-software-supply-chain-security-to-coding-agents-142174.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.23314>)

Author: Aman Sharma, Frank Reyes, Martin Monperrus

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [supply-chain-security](<https://devfeed.tech/topics/supply-chain-security.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [code security](<https://devfeed.tech/topics/code-security.md>), [software composition analysis](<https://devfeed.tech/topics/software-composition-analysis.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>)

Tags: [claude-code](<https://devfeed.tech/tags/claude-code.md>), [claude-opus-5-5](<https://devfeed.tech/tags/claude-opus-5-5.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [deepseek-v4-1-flash](<https://devfeed.tech/tags/deepseek-v4-1-flash.md>), [dependencies](<https://devfeed.tech/tags/dependencies.md>), [github-actions](<https://devfeed.tech/tags/github-actions.md>), [latest-release](<https://devfeed.tech/tags/latest-release.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [registry](<https://devfeed.tech/tags/registry.md>), [security](<https://devfeed.tech/tags/security.md>), [software-supply-chain](<https://devfeed.tech/tags/software-supply-chain.md>), [software-supply-chain-security](<https://devfeed.tech/tags/software-supply-chain-security.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>)

### AI overview

YUL checks each dependency declaration against the package registry and blocks stale or nonexistent versions, telling the coding agent which version to use so it can retry. In 60 tasks across six ecosystems, Claude Code with Claude Opus 5.5 declared stale versions in 29% of runs without YUL, while OpenCode with DeepSeek V4.1 Flash did so in 39%; both did so in every GitHub Actions run. With YUL, agents ended on the latest release in 93% of stale Opus runs and 96% of stale DeepSeek runs. These results suggest registry checks can help coding agents avoid stale dependencies; the supplied abstract does not describe broader security outcomes.

### Source excerpt

arXiv:2611.23314v1 Announce Type: new Abstract: Coding agents choose and declare third-party dependencies on their own. They often declare a stale version that they recall from their training data. A stale dependency exposes the project to known vulnerabilities that the latest release already fixes. We present YUL, a design for coding agent harnesses that ensures that no stale dependencies are ever used. It does so by checking every dependency declaration against the package registry. If a dependency is stale or does not exist, YUL blocks the usage and tells the coding agent which version to use, so the coding agent retries with the fix. We evaluate YUL on 60 tasks across six ecosystems with Claude Code running Claude Opus 5.5 and OpenCode running DeepSeek V4.1 Flash. Without YUL, Opus declares a stale version in 29% of runs and DeepSeek in 39%, and both declare a stale version in every GitHub Actions run. With YUL, the coding agents end on the latest release in 93% of stale Opus runs and in 96% of stale DeepSeek runs. YUL thus brings the best practices of software supply chain security to coding agents.

## VIVID: A Visual Harness for Agentic Video Editing

DevFeed: [VIVID: A Visual Harness for Agentic Video Editing](<https://devfeed.tech/articles/vivid-a-visual-harness-for-agentic-video-editing-137602.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.20688>)

Author: Wenhan Yu, Zhaoxi Zhang, Zhilong Tan, Weikang Li, Lei Sha

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [multimodal-ai](<https://devfeed.tech/topics/multimodal-ai.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [agents](<https://devfeed.tech/tags/agents.md>), [assembly](<https://devfeed.tech/tags/assembly.md>), [audio](<https://devfeed.tech/tags/audio.md>), [audio-analysis](<https://devfeed.tech/tags/audio-analysis.md>), [code](<https://devfeed.tech/tags/code.md>), [code-execution](<https://devfeed.tech/tags/code-execution.md>), [codex](<https://devfeed.tech/tags/codex.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-mm](<https://devfeed.tech/tags/cs-mm.md>), [harness](<https://devfeed.tech/tags/harness.md>), [improvement](<https://devfeed.tech/tags/improvement.md>), [localization](<https://devfeed.tech/tags/localization.md>), [model-training](<https://devfeed.tech/tags/model-training.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [post-production](<https://devfeed.tech/tags/post-production.md>), [repair](<https://devfeed.tech/tags/repair.md>), [tools](<https://devfeed.tech/tags/tools.md>), [transcription](<https://devfeed.tech/tags/transcription.md>), [video](<https://devfeed.tech/tags/video.md>), [video-editing](<https://devfeed.tech/tags/video-editing.md>)

### AI overview

VIVID adds media tools and evidence tracking to native coding agents for video editing, without additional model training. Agents can inspect video, analyze and transcribe audio, retain observations with media versions and timestamps, make edits with their coding and shell tools, and review rendered outputs for revisions. On 100 AgenticVBench tasks across Assembly, Repair, Sequencing, and Repurpose, Codex with VIVID scored 58.6%, 20.2 percentage points above native Codex. The harness also scored 55.1% with Claude Code and 37.1% with OpenCode, each substantially above its native baseline.

### Source excerpt

arXiv:2611.20688v1 Announce Type: new Abstract: Video editing requires agents to make editing decisions grounded in audiovisual observations and precise temporal localization, and to evaluate their effects in rendered outputs. General-purpose coding agents possess strong comprehension, reasoning, and code execution capabilities, while media interaction and evidence management mechanisms tailored to video tasks can further unlock their editing potential. We introduce VIVID, a visual harness that supports multiple native coding agents in video post-production without additional model training. VIVID provides tools for on-demand video inspection, audio analysis, and transcription, and preserves observations alongside their corresponding media versions and temporal coordinates. Agents can actively retrieve these observations as they reason, perform edits using their native coding and shell tools, and inspect rendered outputs to guide subsequent revisions. On AgenticVBench, comprising 100 tasks across Assembly, Repair, Sequencing, and Repurpose, Codex with VIVID achieves an average score of 58.6%, an improvement of 20.2 percentage points over native Codex. The same harness also integrates with Claude Code and OpenCode, achieving average scores of 55.1% and 37.1%, respectively, both substantially exceeding their native baselines. These results highlight VIVID's potential as a reusable visual harness for extending the capabilities of general-purpose coding agents to video editing.

## Use prompt caching with open-weight models on Amazon Bedrock

DevFeed: [Use prompt caching with open-weight models on Amazon Bedrock](<https://devfeed.tech/articles/use-prompt-caching-with-open-weight-models-on-amazon-bedrock-136125.md>)

Original publisher: [Read original article](<https://aws.amazon.com/blogs/machine-learning/use-prompt-caching-with-open-weight-models-on-amazon-bedrock/>)

Author: Dustin Liu

Published: 2026-11-16T15:42:41Z

Content type: tutorial

Language: en

Sources: [Artificial Intelligence](<https://devfeed.tech/sources/artificial-intelligence.md>)

Topics: [Amazon Bedrock](<https://devfeed.tech/topics/amazon-bedrock.md>), [AI-assisted coding](<https://devfeed.tech/topics/ai-assisted-coding.md>), [prompt](<https://devfeed.tech/topics/prompt.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [amazon-bedrock](<https://devfeed.tech/tags/amazon-bedrock.md>), [api](<https://devfeed.tech/tags/api.md>), [best-practices](<https://devfeed.tech/tags/best-practices.md>), [caching](<https://devfeed.tech/tags/caching.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [cost](<https://devfeed.tech/tags/cost.md>), [intermediate-200](<https://devfeed.tech/tags/intermediate-200.md>), [latency](<https://devfeed.tech/tags/latency.md>), [long-context](<https://devfeed.tech/tags/long-context.md>), [open-weight-models](<https://devfeed.tech/tags/open-weight-models.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [prompt-caching](<https://devfeed.tech/tags/prompt-caching.md>), [technical-how-to](<https://devfeed.tech/tags/technical-how-to.md>)

### AI overview

Amazon Bedrock prompt caching can reduce input cost and latency by reusing matching prompt prefixes for open-weight models. Implicit caching works automatically on a best-effort basis; explicit caching uses breakpoints to control which reusable content is stored and avoid writing changing tails that may never be reused. Cache reads cost less than standard input, while some models charge more for writes. The examples show how to configure both modes through the Chat Completions API and verify cache usage in responses, then use implicit caching with OpenCode, which resends its system prompt, tools, and conversation history. Results depend on model and API support, minimum prefix lengths, and cache TTL. The post recommends keeping static content first, checking usage fields and runtime metrics, and keeping frequently reused prefixes warm.

### Source excerpt

Many generative AI applications resend the same long context on every request. Prompt caching lets open-weight models on Amazon Bedrock reuse it to reduce input cost and latency. This post shows implicit and explicit caching through the Chat Completions API and applies caching to an OpenCode coding agent session.

## Programmable Distributions of Semantics-Preserving Variants for Evaluating Code Agents

DevFeed: [Programmable Distributions of Semantics-Preserving Variants for Evaluating Code Agents](<https://devfeed.tech/articles/programmable-distributions-of-semantics-preserving-variants-for-evaluating-code-agents-135001.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2608.18389>)

Author: Hasan Najib Mahmud, Shreya Gupta, Isha Chaudhary, Nathaniel Enis, Ravi Mangal, Gagandeep Singh, Corina Pasareanu

Published: 2026-11-16T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [Refactoring](<https://devfeed.tech/topics/refactoring.md>), [AI-assisted coding](<https://devfeed.tech/topics/ai-assisted-coding.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [agents](<https://devfeed.tech/tags/agents.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-opus](<https://devfeed.tech/tags/claude-opus.md>), [code](<https://devfeed.tech/tags/code.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [distributions](<https://devfeed.tech/tags/distributions.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [frontier-models](<https://devfeed.tech/tags/frontier-models.md>), [kimi](<https://devfeed.tech/tags/kimi.md>), [minimax](<https://devfeed.tech/tags/minimax.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [probabilistic](<https://devfeed.tech/tags/probabilistic.md>), [programming-language](<https://devfeed.tech/tags/programming-language.md>), [qwen](<https://devfeed.tech/tags/qwen.md>), [refactoring](<https://devfeed.tech/tags/refactoring.md>), [reliability](<https://devfeed.tech/tags/reliability.md>), [robustness](<https://devfeed.tech/tags/robustness.md>), [semantics](<https://devfeed.tech/tags/semantics.md>), [swe-bench](<https://devfeed.tech/tags/swe-bench.md>), [swe-bench-pro](<https://devfeed.tech/tags/swe-bench-pro.md>)

### AI overview

The study tests whether code agents can resolve repository issues after surrounding code is rewritten into semantically equivalent variants. It introduces VDL, a probabilistic programming language for defining distributions of variants through randomly applied semantics-preserving transformations. Using a generic distribution, the authors evaluate mini-SWE agent and OpenCode with four underlying models on SWE-bench Verified and SWE-bench Pro. Resolve rates fall by as much as 6.7 percentage points, with significant drops in 6 of 16 configurations. Model robustness rankings do not hold across the two scaffolds, suggesting robustness depends on the scaffold-model pairing. The supplied abstract does not specify individual transformation types or per-configuration results.

### Source excerpt

arXiv:2608.18389v2 Announce Type: replace Abstract: Code agents are increasingly deployed to resolve real software issues, yet their reliability under superficial code variations remains poorly understood. We evaluate whether coding agents that repair repository-level issues remain reliable when the surrounding codebase is rewritten into a semantically equivalent form. We sample semantically equivalent variants of a repository from an explicitly defined distribution and compare the agent's resolve rate on them with that on the original. Different distributions reveal different aspects of an agent, such as its sensitivity to routine refactoring or to misleading code near the location of a fix. We develop VDL, a probabilistic programming language for expressing such distributions, whose programs randomly apply semantics-preserving transformations (SPTs) to a repository. For our evaluation, we use it to define a generic distribution that is not targeted at any particular failure mode. We evaluate two agentic scaffolds (mini-SWE agent and OpenCode), each backed by one of four frontier models (Claude Opus 4.5, Kimi K2.5, MiniMax M2.5, and Qwen 3.6-27B), on instances from SWE-bench Verified and SWE-bench Pro. Resolve rates drop by up to 6.7 percentage points, significantly in 6 of 16 configurations. No ranking of models by robustness holds across scaffolds, revealing a jagged robustness frontier.

## Introducing Kimi K3.1 Pro on Amazon Bedrock

DevFeed: [Introducing Kimi K3.1 Pro on Amazon Bedrock](<https://devfeed.tech/articles/introducing-kimi-k3-1-pro-on-amazon-bedrock-127754.md>)

Original publisher: [Read original article](<https://aws.amazon.com/blogs/machine-learning/introducing-kimi-k3-1-pro-on-amazon-bedrock/>)

Author: Aris Tsakpinis

Published: 2026-11-11T01:07:19Z

Content type: release

Language: en

Sources: [Artificial Intelligence](<https://devfeed.tech/sources/artificial-intelligence.md>)

Topics: [Chat completions endpoints](<https://devfeed.tech/topics/chat-completions-endpoints.md>), [coding assistant](<https://devfeed.tech/topics/coding-assistant.md>), [prompt](<https://devfeed.tech/topics/prompt.md>), [LLM security](<https://devfeed.tech/topics/llm-security.md>), [Amazon CloudWatch](<https://devfeed.tech/topics/amazon-cloudwatch.md>), [amazon](<https://devfeed.tech/topics/amazon.md>), [litellm](<https://devfeed.tech/topics/litellm.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [amazon](<https://devfeed.tech/tags/amazon.md>), [amazon-bedrock](<https://devfeed.tech/tags/amazon-bedrock.md>), [announcements](<https://devfeed.tech/tags/announcements.md>), [apis](<https://devfeed.tech/tags/apis.md>), [bedrock](<https://devfeed.tech/tags/bedrock.md>), [caching](<https://devfeed.tech/tags/caching.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-assistant](<https://devfeed.tech/tags/coding-assistant.md>), [foundational-100](<https://devfeed.tech/tags/foundational-100.md>), [inference](<https://devfeed.tech/tags/inference.md>), [k3](<https://devfeed.tech/tags/k3.md>), [kimi](<https://devfeed.tech/tags/kimi.md>), [kimi-k3](<https://devfeed.tech/tags/kimi-k3.md>), [kimi-k3-1-pro](<https://devfeed.tech/tags/kimi-k3-1-pro.md>), [moonshot](<https://devfeed.tech/tags/moonshot.md>), [moonshot-ai](<https://devfeed.tech/tags/moonshot-ai.md>), [openai](<https://devfeed.tech/tags/openai.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [optimize](<https://devfeed.tech/tags/optimize.md>), [prompt](<https://devfeed.tech/tags/prompt.md>), [prompt-caching](<https://devfeed.tech/tags/prompt-caching.md>)

### AI overview

Kimi K3.1 Pro Highspeed is available on Amazon Bedrock through Chat Completions, Responses, Converse, and Invoke APIs. The article recommends Chat Completions and the lower-cost global inference profile for most workloads; US and India profiles keep processing within those regions. Reduce response time and output tokens on simpler tasks by setting reasoning effort to low or high instead of the max default. Flex costs half the Standard price but can take longer, and on-demand quota is shared across tiers. Explicit prompt caching can reduce latency and input cost for repeated context on Chat Completions, Responses, and Invoke, but not Converse; stable content must reach 256 tokens, and cache writes cost more while reads are discounted. The article also explains connecting OpenCode through Bedrock's Chat Completions provider. AWS data-boundary processing, no sharing with Moonshot AI, and zero data retention are stated.

### Source excerpt

Kimi K3.1 Pro from Moonshot AI is now available on Amazon Bedrock. Learn how to get started through the OpenAI-compatible Chat Completions, Responses, and Converse APIs, optimize inference with explicit prompt caching, and use the model as a coding assistant in OpenCode.

## Trellis: Bringing Compiler-Guided Repair into Agentic Code Generation

DevFeed: [Trellis: Bringing Compiler-Guided Repair into Agentic Code Generation](<https://devfeed.tech/articles/trellis-bringing-compiler-guided-repair-into-agentic-code-generation-113466.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2605.15238>)

Author: Alexander Du, Jianjun Ou, Danyang Zhuo, Matthew Lentz

Published: 2026-11-03T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [AI-assisted coding](<https://devfeed.tech/topics/ai-assisted-coding.md>), [Compilers and Linkers](<https://devfeed.tech/topics/compilers-and-linkers.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [asynchronous](<https://devfeed.tech/tags/asynchronous.md>), [c-c-plus-plus](<https://devfeed.tech/tags/c-c-plus-plus.md>), [c-plus-plus](<https://devfeed.tech/tags/c-plus-plus.md>), [clang](<https://devfeed.tech/tags/clang.md>), [code-generation](<https://devfeed.tech/tags/code-generation.md>), [compilation](<https://devfeed.tech/tags/compilation.md>), [compiler](<https://devfeed.tech/tags/compiler.md>), [costs](<https://devfeed.tech/tags/costs.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-pl](<https://devfeed.tech/tags/cs-pl.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [incremental](<https://devfeed.tech/tags/incremental.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [repair](<https://devfeed.tech/tags/repair.md>), [tasks](<https://devfeed.tech/tags/tasks.md>), [test](<https://devfeed.tech/tags/test.md>), [token-consumption](<https://devfeed.tech/tags/token-consumption.md>), [trellis](<https://devfeed.tech/tags/trellis.md>), [typescript](<https://devfeed.tech/tags/typescript.md>)

### AI overview

Trellis uses incremental compiler feedback to repair coding-agent output, based on the observation that many compile errors are small and local to recent edits. It applies rollback-based search for targeted repairs, with asynchronous checking and checkpointing to control execution costs. The authors adapted Clang and tsc to its checker interface and integrated it into OpenCode. On C/C++ and TypeScript software engineering tasks, Trellis improved compilation rates by up to 13% and test pass rates by up to 11%, with token consumption comparable to baseline OpenCode.

### Source excerpt

arXiv:2605.15238v2 Announce Type: replace Abstract: Coding agents regularly generate programs that do not compile and require substantial tokens to repair. Yet most repairs are small and local to the edit that caused them. We present Trellis, a system that integrates incremental compiler feedback into code generation to improve code correctness while controlling repair costs. Trellis uses rollback-based search for targeted repair and asynchronous checking and checkpointing for efficient execution. We adapt Clang and tsc to Trellis's checker interface and integrate Trellis transparently into OpenCode. On C/C++ and TypeScript software engineering tasks, Trellis improves compilation rate by up to 13% and test pass rate by up to 11% with comparable token consumption to baseline OpenCode.

## Security Impact of Coding Agent's Memory Files

DevFeed: [Security Impact of Coding Agent's Memory Files](<https://devfeed.tech/articles/security-impact-of-coding-agent-s-memory-files-107378.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.38521>)

Author: Thiago Santos de Moura, Selma G\\"uler, Yannic Noller

Published: 2026-10-30T04:00:00Z

Content type: news

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [code security](<https://devfeed.tech/topics/code-security.md>), [Security](<https://devfeed.tech/topics/security.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>)

Tags: [attack-surface](<https://devfeed.tech/tags/attack-surface.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [code-generation](<https://devfeed.tech/tags/code-generation.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [malicious](<https://devfeed.tech/tags/malicious.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [security](<https://devfeed.tech/tags/security.md>), [study](<https://devfeed.tech/tags/study.md>), [vulnerable-code](<https://devfeed.tech/tags/vulnerable-code.md>)

### AI overview

The study finds that malicious project memory files can steer coding agents without changing the developer's prompt, but their measured effect on secure completions was small and inconsistent: rates shifted by −8 to +6 percentage points across the tested agents and variants. Researchers tested Claude Code (Sonnet 5.5) and OpenCode (Qwen 3.8 27B, DeepSeek V4.1 Flash) on 36 SecRepoBench tasks, using qualified and forceful malicious files across 12 CWE classes. Disclosure differed: Claude Code named the file in about two thirds of final messages, while OpenCode mentioned it almost exclusively in its reasoning trace. Flagging a file as an attack did not reliably prevent vulnerable code. The authors call for larger studies to quantify risk across agents and develop defenses.

### Source excerpt

arXiv:2610.38521v1 Announce Type: new Abstract: Coding agents automatically load project memory files such as CLAUDE.md at the start of every session and treat them as trusted project guidance. Anyone who can change such a file, for example through a pull request, can therefore steer the agent without ever touching the developer's prompt. We study memory files as an attack surface for insecure code generation and ask (1) whether a malicious memory file makes agents produce code that passes its tests but is vulnerable, and (2) whether agents disclose to the developer that the file shaped their code. For each of 12 CWE classes, we crafted two malicious memory files, a qualified variant that presents removing defensive checks as engineering best practice and a forceful variant that demands it. We ran Claude Code (Sonnet 5.5) and OpenCode (Qwen 3.8 27B, DeepSeek V4.1 Flash) on 36 SecRepoBench tasks with and without these files. The malicious files changed the rate of secure completions by -8 to +6 percentage points, with no consistent pattern across agents or variants. Agents differ sharply in what they tell the developer about the file. Claude Code names it in about two thirds of its final messages, whereas OpenCode mentions it almost exclusively in its reasoning trace, not in the final message the developer reads. Flagging the file as an attack does not reliably protect the code, as such agents still generated vulnerable code. We outline the larger studies needed to quantify this risk across agents and to design defenses.

## kw-bench: Breaking Frontier Harnesses on Real-World Tasks

DevFeed: [kw-bench: Breaking Frontier Harnesses on Real-World Tasks](<https://devfeed.tech/articles/kw-bench-breaking-frontier-harnesses-on-real-world-tasks-103128.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.36405>)

Author: George Juraj Salapa

Published: 2026-10-28T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [Benchmark](<https://devfeed.tech/topics/benchmark.md>), [Claude Code](<https://devfeed.tech/topics/claude-code.md>), [codex](<https://devfeed.tech/topics/codex.md>)

Tags: [agent-harness](<https://devfeed.tech/tags/agent-harness.md>), [audit](<https://devfeed.tech/tags/audit.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [claude-opus-5-5](<https://devfeed.tech/tags/claude-opus-5-5.md>), [codex](<https://devfeed.tech/tags/codex.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [deepseek](<https://devfeed.tech/tags/deepseek.md>), [deepseek-v4-1-flash](<https://devfeed.tech/tags/deepseek-v4-1-flash.md>), [frontier](<https://devfeed.tech/tags/frontier.md>), [harness](<https://devfeed.tech/tags/harness.md>), [knowledge-work](<https://devfeed.tech/tags/knowledge-work.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [real-world](<https://devfeed.tech/tags/real-world.md>), [tasks](<https://devfeed.tech/tags/tasks.md>)

### AI overview

kw-bench tests investigative knowledge work by asking agents to reconstruct five years of VAT data from undocumented, inconsistent extracts spanning three generations of SAP; grading is deterministic and uses no LLM judge. In a two-hour run with DeepSeek v4.1 Flash, model-agnostic OpenCode and micro-cc scored 0.542 and 0.528, while Claude Code and Codex scored 0 away from their home model. The authors attribute the gap to difficulty retaining weak signals over long sessions. With Claude Opus 5.5, micro-cc and Claude Code scored 0.9986 and 1.000 in 22 and 25 minutes. These results suggest harness choice mattered substantially below the frontier on this task; the abstract does not establish performance across other task families.

### Source excerpt

arXiv:2610.36405v1 Announce Type: new Abstract: Each new generation of frontier model pushes the edge of what is possible. Inside a harness -- a loop, a shell, a filesystem, the ability to write and run its own code -- a model can take on increasingly difficult work. A growing body of work treats the coding-agent harness as infrastructure for knowledge work rather than a coding tool, and a new wave of benchmarks tests it there. Most of them are generative: build a financial model, a memo, a spreadsheet from a brief, judged by a rubric or by another model. Few test the expensive part of knowledge work, which is investigative. In audit, tax, due diligence, regulatory reporting, claims or any dispute over numbers, the answer already exists -- hidden in incomplete and inconsistent evidence spread across systems -- and has to be recovered and defended to the cent. We introduce kw-bench, a benchmark built from real engagements specialists described to us. Its first task family is a five-year VAT reconstruction across three generations of SAP -- legacy ECC, a partial S/4HANA migration and a consultant's redesign -- delivered as raw system extracts: page headers, local number formats, partial link tables, no documentation of how anything connects. Grading is deterministic, with no LLM judge. We ran four harnesses -- Claude Code, Codex, OpenCode and micro-cc -- on the same model, DeepSeek v4.1 Flash, for two hours. The two independent, model-agnostic harnesses delivered (OpenCode 0.542, micro-cc 0.528); the two lab harnesses, away from their home model, delivered nothing (0.00 each). What defeated them was inertia: keeping many weak signals alive across a long session. On a frontier model the task collapses. micro-cc and Claude Code on Claude Opus 5.5 scored 0.9986 and 1.000, in 22 and 25 minutes. A weeks-long engagement now costs minutes and a few dollars; below the frontier, the harness decides how far a model gets.

## Hidden in Plain Sight: an Empirical Study about The Instruction Stacks of Coding Agents

DevFeed: [Hidden in Plain Sight: an Empirical Study about The Instruction Stacks of Coding Agents](<https://devfeed.tech/articles/hidden-in-plain-sight-an-empirical-study-about-the-instruction-stacks-of-coding-agents-103656.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.37332>)

Author: Felipe de Sant'Anna Paix\\~ao, Dhyego Tavares, Eduardo Almeida, Iftekhar Ahmed

Published: 2026-10-28T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [AI-assisted coding](<https://devfeed.tech/topics/ai-assisted-coding.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [claude-code](<https://devfeed.tech/tags/claude-code.md>), [codex](<https://devfeed.tech/tags/codex.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [coding-style](<https://devfeed.tech/tags/coding-style.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [gemini-cli](<https://devfeed.tech/tags/gemini-cli.md>), [instructions](<https://devfeed.tech/tags/instructions.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [qwen](<https://devfeed.tech/tags/qwen.md>), [research](<https://devfeed.tech/tags/research.md>), [stack](<https://devfeed.tech/tags/stack.md>), [tool-use](<https://devfeed.tech/tags/tool-use.md>)

### AI overview

Across 1,886 releases of Claude Code, Codex, Gemini CLI, Qwen Code, and OpenCode, the study finds that coding agents' first model requests include evolving instruction stacks--tools and environmental context as well as prompts--that are rarely visible or fully documented. Tool definitions dominate three agents' latest requests, and concentrated revisions sometimes change guidance on tool use, attribution, and coding style, complicating analysis of agent-authored code. Only 48% of examined change units directly explained revision motivations. The authors recommend documenting captured instruction stacks alongside releases, models, and user prompts, and preserving revision rationales.

### Source excerpt

arXiv:2610.37332v1 Announce Type: new Abstract: Coding agents send their models more than the user's prompt: they add instructions, tool definitions, and environmental context that form an instruction stack. This stack guides execution and changes across releases, yet remains largely unexposed to users and insufficiently characterized in research. We present a longitudinal, cross-agent study of the instruction stacks captured in the first model requests of 1,886 releases of Claude Code, Codex, Gemini CLI, Qwen Code, and OpenCode. Under controlled conditions, we analyze their composition and evolution and trace changes in the four open-source agents to documented motivations. Our findings show why examining the system prompt or reporting the model alone provides an incomplete account of agent inputs. Tool definitions dominate three agents' latest requests, while occasional concentrated revisions alter instructions governing tool use, attribution, and coding style. Such revisions can complicate how researchers identify agent-authored contributions and interpret generated code. We also found that only 48% of examined change units directly explain the motivation for stack revisions. Across agents, shared capabilities emerge despite increasingly distinct wording. These findings expose an evolving, incompletely documented layer of agent design. The findings motivate documenting captured instruction stacks alongside agent releases, models, and user prompts, and preserving the rationale behind their revisions.

## Vercel Plugin now available in OpenCode

DevFeed: [Vercel Plugin now available in OpenCode](<https://devfeed.tech/articles/vercel-plugin-now-available-in-opencode-102771.md>)

Original publisher: [Read original article](<https://vercel.com/changelog/vercel-plugin-now-available-in-opencode>)

Author: Vincent Voyer

Published: 2026-10-27T00:00:00Z

Content type: release

Language: en

Sources: [Vercel News](<https://devfeed.tech/sources/vercel-news.md>)

Topics: [Vercel](<https://devfeed.tech/topics/vercel.md>), [Cloud Architecture](<https://devfeed.tech/topics/cloud-architecture.md>), [Environment Variables](<https://devfeed.tech/topics/environment-variables.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-application](<https://devfeed.tech/tags/ai-application.md>), [ai-sdk](<https://devfeed.tech/tags/ai-sdk.md>), [apis](<https://devfeed.tech/tags/apis.md>), [deploy](<https://devfeed.tech/tags/deploy.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [design](<https://devfeed.tech/tags/design.md>), [documentation](<https://devfeed.tech/tags/documentation.md>), [environment-variables](<https://devfeed.tech/tags/environment-variables.md>), [functions](<https://devfeed.tech/tags/functions.md>), [guidance](<https://devfeed.tech/tags/guidance.md>), [install](<https://devfeed.tech/tags/install.md>), [learn](<https://devfeed.tech/tags/learn.md>), [manage](<https://devfeed.tech/tags/manage.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-server](<https://devfeed.tech/tags/mcp-server.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [plugin](<https://devfeed.tech/tags/plugin.md>), [vercel](<https://devfeed.tech/tags/vercel.md>)

### AI overview

The Vercel Plugin for OpenCode provides on-demand guidance on Vercel APIs and recommended patterns, including skills for AI SDK, Vercel Functions, and Workflow, plus specialist agents for deployment, performance optimization, and AI application design. It requires OpenCode 2. Install it with `opencode plugin add github:vercel/vercel-plugin`, then run `/mcps`, select `vercel`, and sign in to connect the Vercel MCP server. Use `/vercel:deploy` to deploy projects and `/vercel:env` to manage environment variables.

### Source excerpt

The Vercel Plugin is now available in OpenCode. The plugin provides on-demand guidance on Vercel APIs and recommended patterns, with skills for AI SDK, Vercel Functions, Workflow, and more. It also includes specialist agents for deployment, performance optimization, and AI application design. Use /vercel:deploy to deploy projects and /vercel:env to manage environment variables. The plugin requires OpenCode 2. To install it, run this command in your terminal: Then run /mcps in OpenCode, select vercel, and sign in to connect the Vercel MCP server to your account. Learn more in the Vercel Plugin documentation. Read more

## HarnessAgent now supports image inputs for all harness adapters

DevFeed: [HarnessAgent now supports image inputs for all harness adapters](<https://devfeed.tech/articles/harnessagent-now-supports-image-inputs-for-all-harness-adapters-102438.md>)

Original publisher: [Read original article](<https://vercel.com/changelog/harness-agent-image-inputs>)

Author: Josh Singh

Published: 2026-10-27T00:00:00Z

Content type: release

Language: en

Sources: [Vercel News](<https://devfeed.tech/sources/vercel-news.md>)

Topics: [vlm](<https://devfeed.tech/topics/vlm.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>)

Tags: [adapters](<https://devfeed.tech/tags/adapters.md>), [ai-sdk](<https://devfeed.tech/tags/ai-sdk.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [cline](<https://devfeed.tech/tags/cline.md>), [codex](<https://devfeed.tech/tags/codex.md>), [cursor](<https://devfeed.tech/tags/cursor.md>), [github-copilot](<https://devfeed.tech/tags/github-copilot.md>), [grok-build](<https://devfeed.tech/tags/grok-build.md>), [harness](<https://devfeed.tech/tags/harness.md>), [image](<https://devfeed.tech/tags/image.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [pi](<https://devfeed.tech/tags/pi.md>), [vision](<https://devfeed.tech/tags/vision.md>)

### AI overview

HarnessAgent accepts images alongside text in a standard AI SDK message: use a `content` array with `{ type: 'text' }` and `{ type: 'file' }` parts. The file part includes an image media type and image data; the example reads a PNG and asks the Codex harness to describe it. Images are processed by each harness's underlying vision model. The feature is supported by the listed adapters: Claude Code, Cline, Codex, Cursor, Deep Agents, fx, GitHub Copilot, Grok Build, OpenCode, and Pi.

### Source excerpt

The AI SDK HarnessAgent layer now accepts image inputs across every harness adapter. Pass images alongside text in the standard AI SDK message format: a content array mixing { type: 'text' } and { type: 'file' } parts. Image input is processed by the harnesses' underlying vision models and supported across all available adapters: Claude Code, Cline, Codex, Cursor, Deep Agents, fx, GitHub Copilot, Grok Build, OpenCode, Pi. Read the HarnessAgent image inputs documentation to get started. Read more

## When Agents Fail: Model or Harness--Who's to Blame?

DevFeed: [When Agents Fail: Model or Harness--Who's to Blame?](<https://devfeed.tech/articles/when-agents-fail-model-or-harness-who-s-to-blame-100016.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.32483>)

Author: Shao Zhang, Hangfan Zhang, Yuanyuan Shi, Hao Li, Min-Le Su, Kangcong Li, Jingping Wang, Chunjiang Mu, Chen Zhang, Yang Chen, Bo Zhang, Shuyue Hu

Published: 2026-10-26T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [harvestbench](<https://devfeed.tech/topics/harvestbench.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [benchmark-performance](<https://devfeed.tech/tags/benchmark-performance.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [context](<https://devfeed.tech/tags/context.md>), [context-management](<https://devfeed.tech/tags/context-management.md>), [covering](<https://devfeed.tech/tags/covering.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [deepseek](<https://devfeed.tech/tags/deepseek.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [execution](<https://devfeed.tech/tags/execution.md>), [failed](<https://devfeed.tech/tags/failed.md>), [harness](<https://devfeed.tech/tags/harness.md>), [load](<https://devfeed.tech/tags/load.md>), [long-context](<https://devfeed.tech/tags/long-context.md>), [model](<https://devfeed.tech/tags/model.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [pairing](<https://devfeed.tech/tags/pairing.md>), [rankings](<https://devfeed.tech/tags/rankings.md>), [tool-use](<https://devfeed.tech/tags/tool-use.md>), [training](<https://devfeed.tech/tags/training.md>)

### AI overview

Agent benchmark results reflect the combined behavior of models and execution harnesses, so the study evaluates their interactions. It proposes five measures: Valid Completion Rate (normal completion with an evaluable result, regardless of correctness), Model-Harness Fit, Cross-Model Fit, pairing-specific Interaction Residuals, and Token Load, including failed attempts. Across four general-purpose harnesses, four core models, two fixed native systems, and six benchmarks, 111 settings produced 52,168 trajectories. Rankings varied by model and benchmark; task scores and valid completion could favor different pairings, and similar scores could mask different execution profiles. OpenCode ranked first in cross-model performance, while DeepSeek V4 Flash Preview had consistently high Mean VCR across harnesses on the shared four-benchmark panel. The authors recommend interaction-focused evaluation, transferable harness-use training for models, and harness designs adapted to model and task.

### Source excerpt

arXiv:2610.32483v1 Announce Type: new Abstract: LLM agents combine models with execution harnesses, so agentic benchmark performance reflects their joint behavior. We argue that harnesses should therefore be evaluated through their interactions with underlying models. We develop a framework combining five complementary measures: (i) Valid Completion Rate (VCR), how reliably agents finish normally and deliver evaluable results, irrespective of task correctness; (ii) Model-Harness Fit, how well a harness supports task achievement through valid completion for a given model; (iii) Cross-Model Fit, a harness's overall suitability across evaluated models; (iv) Interaction Residuals, pairing-specific execution strengths and weaknesses hidden by component-level summaries; and (v) Token Load, the token demands of agent execution, including failed attempts. We evaluate four general-purpose harnesses (OpenCode, Codex CLI, Pi CLI, and DeepAgents) with four core models, plus two fixed native model-harness systems, on six benchmarks covering context management, tool use, long-context planning, multimodal interaction, and artifact handoff. Including native references and stress conditions, 111 evaluation settings yield 52,168 trajectories. Harness rankings vary across models and benchmarks; task performance and valid completion can favor different pairings, and similar task scores can accompany distinct execution profiles. OpenCode ranks first in cross-model performance, while DeepSeek V4 Flash Preview maintains consistently high Mean VCR across harnesses on the shared four-benchmark panel. These findings support evaluating model-harness interactions, training transferable harness-use capabilities in models, and harness designs adapted to the model and task.

## Benchmarks Are Surprisingly Compaction-Friendly

DevFeed: [Benchmarks Are Surprisingly Compaction-Friendly](<https://devfeed.tech/articles/benchmarks-are-surprisingly-compaction-friendly-99508.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.31778>)

Author: Jeanne Salle, Hardik Bhatnagar, Matthias Bethge, Sahar Abdelnabi, Maksym Andriushchenko

Published: 2026-10-26T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [context window](<https://devfeed.tech/topics/context-window.md>), [LLM evaluation / benchmarking](<https://devfeed.tech/topics/llm-evaluation-benchmarking.md>), [Ground truth / benchmark quality](<https://devfeed.tech/topics/ground-truth-benchmark-quality.md>), [long-running](<https://devfeed.tech/topics/long-running.md>), [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-tasks](<https://devfeed.tech/tags/agentic-tasks.md>), [agents](<https://devfeed.tech/tags/agents.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [compaction](<https://devfeed.tech/tags/compaction.md>), [context](<https://devfeed.tech/tags/context.md>), [context-window](<https://devfeed.tech/tags/context-window.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [friendly](<https://devfeed.tech/tags/friendly.md>), [frontier-models](<https://devfeed.tech/tags/frontier-models.md>), [gpt](<https://devfeed.tech/tags/gpt.md>), [gpt-5](<https://devfeed.tech/tags/gpt-5.md>), [llm](<https://devfeed.tech/tags/llm.md>), [llm-agents](<https://devfeed.tech/tags/llm-agents.md>), [open-weight-models](<https://devfeed.tech/tags/open-weight-models.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [performance](<https://devfeed.tech/tags/performance.md>)

### AI overview

In tests using OpenCode, five context-compaction strategies, three frontier models and three long-horizon benchmarks, compaction usually had little measurable effect on the two models with 1M-token context windows. GPT-5.3-Codex, with the smallest window, performed worse at lower compaction thresholds on PaperBench and less clearly on PostTrainBench. No tested strategy beat default self-summarization; discarding more content cost some models up to 20 points while leaving others unchanged. Yet summaries could silently drop constraints: the first compaction lost about a third of ProgramBench's anti-cheating rules on average, with losses growing as compactions accumulated. In SWE-chat sessions, instructions about how the agent should work or what to avoid were dropped about six times more often than build specifications. The authors argue these benchmarks are compaction-friendly because objectives stay fixed and progress remains on disk, making most discarded information recoverable--except constraints recorded nowhere else.

### Source excerpt

arXiv:2610.31778v1 Announce Type: new Abstract: Long horizon agentic tasks are now pushing LLM agents past their context window. To continue the task, the agent scaffold has to free up the context window, a process known as context compaction. Compaction is generally lossy by design: agents can silently lose task constraints, instructions, or their reasoning behind earlier decisions. Although compaction is a standard component of coding agent scaffolds, it has not been studied systematically with frontier models on realistic, long horizon agentic tasks. Previous work evaluate smaller open weight models on toy or synthetic trajectories. We fix the agent scaffold (OpenCode), implement five compaction strategies inside it, and evaluate three frontier models on three long horizon benchmarks (PaperBench, ProgramBench, PostTrainBench). For the two 1M context models, the compaction threshold does not measurably affect the task performance, whereas GPT-5.3-Codex, the model with the smallest context window, performs worse at lower thresholds on PaperBench and, less clearly, on PostTrainBench. No compaction method outperforms default self-summarization, and methods that discard more content cost some models up to 20 points while leaving other models the same. Compaction can, however, silently remove task constraints: on ProgramBench, the first compaction summary already drops about one third of the task's anticheating rules on average, and this rises monotonically with the number of compactions. On real user-agent sessions from SWE-chat, instructions about how the agent should work and what it should not do are dropped around 6 times more than specifications of what to build. We argue that current benchmarks are compaction-friendly: i.e., the objective is fixed and the agent's progress stays on disk, so most of what compaction discards can be recovered, except constraints that don't exist anywhere else.

## Usage metrics API adds usage from tools like OpenCode

DevFeed: [Usage metrics API adds usage from tools like OpenCode](<https://devfeed.tech/articles/usage-metrics-api-adds-usage-from-tools-like-opencode-98584.md>)

Original publisher: [Read original article](<https://github.blog/changelog/2026-10-23-usage-metrics-api-adds-usage-from-tools-like-opencode>)

Author: Allison

Published: 2026-10-23T23:42:03Z

Content type: release

Language: en

Sources: [GitHub Changelog](<https://devfeed.tech/sources/github-changelog.md>)

Topics: [API](<https://devfeed.tech/topics/api.md>), [Chat completions endpoints](<https://devfeed.tech/topics/chat-completions-endpoints.md>)

Tags: [account-management](<https://devfeed.tech/tags/account-management.md>), [api](<https://devfeed.tech/tags/api.md>), [copilot](<https://devfeed.tech/tags/copilot.md>), [enterprise-management-tools](<https://devfeed.tech/tags/enterprise-management-tools.md>), [improvement](<https://devfeed.tech/tags/improvement.md>), [metrics](<https://devfeed.tech/tags/metrics.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [tools](<https://devfeed.tech/tags/tools.md>)

### AI overview

The Copilot usage metrics API adds `totals_by_client` to enterprise, organization, and user reports covering 1-day and 28-day periods. It groups successful model calls by the client tool and reports request counts, distinct sessions, token usage, and per-model breakdowns, including whether model selection was user-, auto-, or product-selected. This captures Copilot model traffic from tools such as OpenCode that may not send editor telemetry, and differs from agent app activity, which counts jobs started by agent apps. Access requires the Copilot usage metrics policy to be enabled and an eligible owner, billing manager, or role with `View Copilot Metrics`. Treat client labels as best-effort integration IDs; request counts measure model calls, not prompts. Code review models appear as `unknown`, and some selection modes or unrecognized clients are grouped as `other`. The addition is backward compatible, with an empty array when there is no usage.

### Source excerpt

The Copilot usage metrics API now reports model usage from more of the tools your developers use with Copilot, including tools like OpenCode that weren't previously captured in your reports.... The post Usage metrics API adds usage from tools like OpenCode appeared first on The GitHub Blog.

## OpenCode Flaw Lets Malicious MCP Servers Execute Code During Sign-In on Windows and WSL

DevFeed: [OpenCode Flaw Lets Malicious MCP Servers Execute Code During Sign-In on Windows and WSL](<https://devfeed.tech/articles/opencode-flaw-lets-malicious-mcp-servers-execute-code-during-sign-in-on-windows-and-wsl-98287.md>)

Original publisher: [Read original article](<https://thehackernews.com/2026/10/opencode-flaw-lets-malicious-mcp.html>)

Author: The Hacker News

Published: 2026-10-22T10:44:05Z

Content type: news

Language: en

Sources: [Cyber Newsroom](<https://devfeed.tech/sources/cyber-newsroom.md>)

Topics: [code security](<https://devfeed.tech/topics/code-security.md>), [Windows Subsystem for Linux](<https://devfeed.tech/topics/wsl.md>), [Windows](<https://devfeed.tech/topics/windows.md>), [MSP MCP](<https://devfeed.tech/topics/msp-mcp.md>)

Tags: [ai-coding-agent](<https://devfeed.tech/tags/ai-coding-agent.md>), [application-security](<https://devfeed.tech/tags/application-security.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [attack](<https://devfeed.tech/tags/attack.md>), [code](<https://devfeed.tech/tags/code.md>), [code-execution](<https://devfeed.tech/tags/code-execution.md>), [command-injection](<https://devfeed.tech/tags/command-injection.md>), [cve](<https://devfeed.tech/tags/cve.md>), [malicious](<https://devfeed.tech/tags/malicious.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-servers](<https://devfeed.tech/tags/mcp-servers.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [sign-in](<https://devfeed.tech/tags/sign-in.md>), [vulnerability](<https://devfeed.tech/tags/vulnerability.md>), [windows](<https://devfeed.tech/tags/windows.md>), [wsl](<https://devfeed.tech/tags/wsl.md>)

### AI overview

A malicious or hijacked remote MCP server could execute commands as the developer when they initiated sign-in in OpenCode on Windows or WSL. The server supplied an OAuth authorization URL that OpenCode passed unescaped into a PowerShell command used to open the browser. Merely configuring a server or opening a session did not trigger the flaw; the developer had to start MCP sign-in, and the attacker needed control of a server the developer had added. OpenCode fixed the issue in version 1.18.33; versions 1.0.137-1.18.32 are reported affected. macOS and Linux are unaffected. The flaw has no CVE or vendor advisory, so scanners may not identify vulnerable installs. Upgrade to 1.18.33 or later; if that is not possible, remove unvetted remote MCP servers and check configured servers with `opencode mcp list`. The researcher reported no known exploitation.

### Source excerpt

OpenCode Flaw Lets Malicious MCP Servers Execute Code During Sign-In on Windows and WSL A flaw in the OpenCode AI coding agent allowed a malicious or hijacked MCP server to run code on a developer's Windows or WSL machine when the developer signed in. OpenCode connects to MCP (Model Cont ... Read more Published Date: Oct 22, 2026 (1 day, 8 hours ago) Vulnerabilities has been mentioned in this article. CVE-2025-6514

## Compaction Laundering: Provenance Destruction in Context-Compacting Coding Agents

DevFeed: [Compaction Laundering: Provenance Destruction in Context-Compacting Coding Agents](<https://devfeed.tech/articles/compaction-laundering-provenance-destruction-in-context-compacting-coding-agents-94543.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.29869>)

Author: Arshan Dabirsiaghi (Pixee)

Published: 2026-10-22T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [AI Context Management](<https://devfeed.tech/topics/ai-context-management.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [AI software supply chain security](<https://devfeed.tech/topics/ai-software-supply-chain-security.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [context window](<https://devfeed.tech/topics/context-window.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [aider](<https://devfeed.tech/tags/aider.md>), [attack](<https://devfeed.tech/tags/attack.md>), [backdoor](<https://devfeed.tech/tags/backdoor.md>), [cline](<https://devfeed.tech/tags/cline.md>), [code-review](<https://devfeed.tech/tags/code-review.md>), [codex](<https://devfeed.tech/tags/codex.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [compaction](<https://devfeed.tech/tags/compaction.md>), [context](<https://devfeed.tech/tags/context.md>), [context-window](<https://devfeed.tech/tags/context-window.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [curl](<https://devfeed.tech/tags/curl.md>), [gemini-cli](<https://devfeed.tech/tags/gemini-cli.md>), [harness](<https://devfeed.tech/tags/harness.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [pi](<https://devfeed.tech/tags/pi.md>), [provenance](<https://devfeed.tech/tags/provenance.md>), [qwen](<https://devfeed.tech/tags/qwen.md>), [sandbox](<https://devfeed.tech/tags/sandbox.md>)

### AI overview

The reported vulnerability arises when coding-agent harnesses flatten role-typed messages into plain text labels before summarization: an attacker can forge a user turn that survives context compaction as a genuine instruction. The abstract reports forged approvals persisted 87-100% of the time in five flattening harnesses (pi, opencode, Cline, Aider, OpenHands), versus 0% in four role-typed harnesses (Codex, Gemini CLI, Qwen Code, Continue). It describes resulting tool execution and steering an agent away from a designated file, potentially hiding a planted backdoor. A measured code-level provenance check caught 77% of laundering while demoting 2% of legitimate summary content. The proposed serializer fix escapes role labels before flattening; role-typed harnesses preserve provenance. These findings are from the supplied abstract and excerpt.

### Source excerpt

arXiv:2610.29869v1 Announce Type: new Abstract: Coding agents that run long enough must compact their context, summarizing the conversation, discarding the original, and carrying the summary forward as memory. We show that in popular open-source coding agents this step is an exploitable trust boundary. Harnesses that flatten the typed conversation into a single plain string, marking each message with a text label such as [User]:, destroy message provenance in code, before the summarizer runs. An attacker who controls anything the agent reads can forge a user turn that survives compaction as a genuine instruction. Across five flattening harnesses (pi, opencode, Cline, Aider, OpenHands) the forged approval launders into the agent's memory 87-100% of the time, up to a ~908,000-token context window; across four role-typed harnesses (Codex, Gemini CLI, Qwen Code, Continue) the identical attack succeeds 0% of the time. We follow the attack to executed consequences. A laundered standing approval drives real tool-running agents to execute attacker-controlled curl | sh in a sandbox, and end to end in the live pi TUI at ~50% against a 0% paired clean arm. The same channel reliably steers an agent's plan away from a file an attacker designates off-limits, which can blind a code review to a planted backdoor. Because the provenance is destroyed deterministically in harness code, no prompt-level defense can restore it; we measure a code-level provenance check that catches 77% of laundering while demoting 2% of legitimate summary content. The defect also has a straightforward fix at the serializer, escaping the role labels before flattening; the role-typed harnesses avoid it entirely by keeping messages role-typed.

## Vercel Open-Sources durable-harness for Crash-Proof Coding Agents

DevFeed: [Vercel Open-Sources durable-harness for Crash-Proof Coding Agents](<https://devfeed.tech/articles/vercel-open-sources-durable-harness-for-crash-proof-coding-agents-94009.md>)

Original publisher: [Read original article](<https://www.marktechpost.com/2026/10/21/vercel-durable-harness-crash-proof-coding-agents/>)

Author: Michal Sutter

Published: 2026-10-21T22:15:39Z

Content type: article

Language: en

Sources: [MarkTechPost](<https://devfeed.tech/sources/marktechpost.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [Persistence](<https://devfeed.tech/topics/persistence.md>), [Vercel](<https://devfeed.tech/topics/vercel.md>)

Tags: [adapters](<https://devfeed.tech/tags/adapters.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai-shorts](<https://devfeed.tech/tags/ai-shorts.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [codex](<https://devfeed.tech/tags/codex.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [crash](<https://devfeed.tech/tags/crash.md>), [crashes](<https://devfeed.tech/tags/crashes.md>), [editors-pick](<https://devfeed.tech/tags/editors-pick.md>), [for-devs](<https://devfeed.tech/tags/for-devs.md>), [harness](<https://devfeed.tech/tags/harness.md>), [laptop](<https://devfeed.tech/tags/laptop.md>), [logs](<https://devfeed.tech/tags/logs.md>), [model](<https://devfeed.tech/tags/model.md>), [new-releases](<https://devfeed.tech/tags/new-releases.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [package](<https://devfeed.tech/tags/package.md>), [pi](<https://devfeed.tech/tags/pi.md>), [postgres](<https://devfeed.tech/tags/postgres.md>), [sessions](<https://devfeed.tech/tags/sessions.md>), [staff](<https://devfeed.tech/tags/staff.md>), [tech-news](<https://devfeed.tech/tags/tech-news.md>), [technology](<https://devfeed.tech/tags/technology.md>), [tool](<https://devfeed.tech/tags/tool.md>), [vercel](<https://devfeed.tech/tags/vercel.md>)

### AI overview

Vercel's experimental durable-harness logs model and tool call intents and outcomes, then resumes a turn from that log if its function crashes. Completed calls are read back rather than rerun; tools can be configured for safe replay or never to replay. Harness session files are mirrored too. Pi and OpenCode can run in the function, while Claude Code and Codex require a Vercel Sandbox. Workspaces include a networkless virtual filesystem, temporary in-memory scratch files, or a Sandbox with Drive storage. The package can run on Vercel, locally, or self-hosted with Postgres on Node.js 22+; the excerpt also lists MongoDB, Redis, and Turso. It is version 0.0.x, and Vercel has published no performance benchmarks; recovery claims are self-reported.

### Source excerpt

Vercel released durable-harness, an open-source package that logs every model and tool call so Pi, OpenCode, Claude Code and Codex sessions resume after crashes, timeouts and redeploys on Vercel, Postgres or a laptop. The post Vercel Open-Sources durable-harness for Crash-Proof Coding Agents appeared first on MarkTechPost.

## OpenCode hack exposes API keys amid AI distillation attacks

DevFeed: [OpenCode hack exposes API keys amid AI distillation attacks](<https://devfeed.tech/articles/opencode-hack-exposes-api-keys-amid-ai-distillation-attacks-90455.md>)

Original publisher: [Read original article](<https://www.bleepingcomputer.com/news/security/opencode-hack-exposes-api-keys-amid-ai-distillation-attacks/>)

Author: Bill Toulas

Published: 2026-10-19T17:27:15Z

Content type: news

Language: en

Sources: [BleepingComputer](<https://devfeed.tech/sources/bleepingcomputer.md>)

Topics: [Secrets Management](<https://devfeed.tech/topics/secrets-management.md>), [API keys](<https://devfeed.tech/topics/api-keys.md>), [token jacking](<https://devfeed.tech/topics/token-jacking.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-coding](<https://devfeed.tech/tags/ai-coding.md>), [ai-cybersecurity](<https://devfeed.tech/tags/ai-cybersecurity.md>), [api-key](<https://devfeed.tech/tags/api-key.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [breach](<https://devfeed.tech/tags/breach.md>), [charges](<https://devfeed.tech/tags/charges.md>), [computer-help](<https://devfeed.tech/tags/computer-help.md>), [computer-security](<https://devfeed.tech/tags/computer-security.md>), [computers](<https://devfeed.tech/tags/computers.md>), [credentials](<https://devfeed.tech/tags/credentials.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [developers](<https://devfeed.tech/tags/developers.md>), [infosec](<https://devfeed.tech/tags/infosec.md>), [infosec-computer-security](<https://devfeed.tech/tags/infosec-computer-security.md>), [initial-access](<https://devfeed.tech/tags/initial-access.md>), [linux](<https://devfeed.tech/tags/linux.md>), [mac](<https://devfeed.tech/tags/mac.md>), [malware](<https://devfeed.tech/tags/malware.md>), [malware-removal](<https://devfeed.tech/tags/malware-removal.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [private-key](<https://devfeed.tech/tags/private-key.md>), [security](<https://devfeed.tech/tags/security.md>), [security-breach](<https://devfeed.tech/tags/security-breach.md>), [spyware](<https://devfeed.tech/tags/spyware.md>), [support](<https://devfeed.tech/tags/support.md>), [tech-support](<https://devfeed.tech/tags/tech-support.md>), [technical-support](<https://devfeed.tech/tags/technical-support.md>), [virus](<https://devfeed.tech/tags/virus.md>), [virus-removal](<https://devfeed.tech/tags/virus-removal.md>), [windows](<https://devfeed.tech/tags/windows.md>)

### AI overview

OpenCode reported that a breach exposed API keys some users had stored for external AI providers. The incident affected 1.4% of OpenCode Console users and led to reports of unauthorized charges and suspended Google Cloud projects. The article says attackers used AI agents in a large-scale distillation attack, while OpenCode worked with Google, OpenAI, and Anthropic to address the incident.

### Source excerpt

The open-source AI coding project OpenCode has notified a small number of users that API keys they used with external providers were exposed in a security breach. [...]

## Making every harness durable

DevFeed: [Making every harness durable](<https://devfeed.tech/articles/making-every-harness-durable-91278.md>)

Original publisher: [Read original article](<https://vercel.com/blog/making-every-harness-durable>)

Author: Eric Dodds

Published: 2026-10-19T13:00:00Z

Content type: release

Language: en

Sources: [Vercel News](<https://devfeed.tech/sources/vercel-news.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [Multi-step AI processes](<https://devfeed.tech/topics/multi-step-ai-processes.md>), [Persistence](<https://devfeed.tech/topics/persistence.md>), [Cloud](<https://devfeed.tech/topics/cloud.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-workflows](<https://devfeed.tech/tags/agent-workflows.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [code](<https://devfeed.tech/tags/code.md>), [codex](<https://devfeed.tech/tags/codex.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [crash](<https://devfeed.tech/tags/crash.md>), [credentials](<https://devfeed.tech/tags/credentials.md>), [database](<https://devfeed.tech/tags/database.md>), [durability](<https://devfeed.tech/tags/durability.md>), [eve](<https://devfeed.tech/tags/eve.md>), [fx](<https://devfeed.tech/tags/fx.md>), [harness](<https://devfeed.tech/tags/harness.md>), [network](<https://devfeed.tech/tags/network.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [pi](<https://devfeed.tech/tags/pi.md>), [postgres](<https://devfeed.tech/tags/postgres.md>)

### AI overview

`durable-harness` makes supported coding-agent harnesses resume after crashes or timeouts by logging each model and tool call's intent and result, with a queue handing interrupted turns to a new function. Completed calls are read from the log; interrupted tools marked `safe` can rerun, while other tools are presented to the model as possibly executed so it can decide whether to check. It supports Vercel, local development, and self-hosted Worlds including Postgres, MongoDB, Redis, or Turso. Workspace choices range from logged virtual files for short tasks to a persistent Vercel Sandbox for larger or longer-running work. Claude Code and Codex require a Vercel Sandbox for their built-in shell and filesystem tools; Pi and OpenCode have other workspace options. The excerpt says credentials for database and API work can stay in functions rather than entering a sandbox.

### Source excerpt

Developers are familiar with running coding agents locally through harnesses like Pi, Claude Code, and others. They are reliable daily drivers for generating code, but keeping agents running around the clock requires your computer to be on and awake, which led to the laptop lid memes and Mac mini demand spike. Running harnesses in the cloud Harness workloads are moving to the cloud. Individual developers are deploying always-on coding agent workflows, projects like OpenWork and HanaAgent use open-source harnesses as the engines behind the apps, and engineers are building their own harness products. These harnesses are designed for long-lived processes that keep the session in memory, and agent tasks commonly run for minutes or hours. Locally, a harness has direct, continuous access to the workspace and session history on the same disk, so resuming interrupted work only requires a simple prompt. In the cloud, you can replicate that setup with a VM and persistent storage, but the VM doesn't automatically record which model and tool calls completed or resume interrupted turns after a crash. A failed turn or network error means lost work. Also, not every agent task executed through a harness needs a sandbox. Database queries and API calls can run securely in functions, and that also means required credentials never have to enter the sandbox. Making every harness durable We built eve and fx to run durably in the cloud out of the box, and today we are introducing durable-harness, which extends that durability to any harness, including Pi, OpenCode, Codex, and Claude Code. It uses the same machinery that powers Workflow SDK to survive crashes and function timeouts. With durable-harness, you can: Run any of these harnesses durably on Vercel Self-host a harness with your own Postgres instance Make your local harness workflow durable Build your own durable harness Sessions that survive a crash mid-turn Take the example of a coding agent adding a pricing page to an existing ap

## OpenCode vs. Claude Code: Which AI coding tool is best? \[2026\]

DevFeed: [OpenCode vs. Claude Code: Which AI coding tool is best? \[2026\]](<https://devfeed.tech/articles/opencode-vs-claude-code-which-ai-coding-tool-is-best-2026-90458.md>)

Original publisher: [Read original article](<https://zapier.com/blog/opencode-vs-claude-code>)

Author: Miguel Rebelo

Published: 2026-10-19T04:00:00Z

Content type: comparison

Language: en

Sources: [Zapier](<https://devfeed.tech/sources/zapier.md>)

Topics: [Claude Code](<https://devfeed.tech/topics/claude-code.md>), [coding assistant](<https://devfeed.tech/topics/coding-assistant.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [Claude](<https://devfeed.tech/topics/claude.md>), [anthropic](<https://devfeed.tech/topics/anthropic.md>), [Multi-step AI processes](<https://devfeed.tech/topics/multi-step-ai-processes.md>), [Vibe coding](<https://devfeed.tech/topics/vibe-coding.md>), [remote-development](<https://devfeed.tech/topics/remote-development.md>), [Visual Studio Code](<https://devfeed.tech/topics/visual-studio-code.md>)

Tags: [ai-coding](<https://devfeed.tech/tags/ai-coding.md>), [ai-coding-tools](<https://devfeed.tech/tags/ai-coding-tools.md>), [app-picks](<https://devfeed.tech/tags/app-picks.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [opencode](<https://devfeed.tech/tags/opencode.md>)

### AI overview

Claude Code offers a simpler, more guarded experience with Claude models and subscription-bundled usage; OpenCode is a free, open-source harness that supports many model providers but shifts setup, permissions, and billing management to the user. The article recommends Claude Code for people who want straightforward setup and stronger default protections, and OpenCode for developers who value model choice, local operation, or control over spending and are willing to configure it. OpenCode runs actions without permission prompts by default and has no built-in classifier or sandbox equivalent described here; the article suggests configuring JSON permissions and using a virtual machine. Its free models may not suit complex work, and serious use requires paid model access. Claude Code's subscription includes usage but has rolling and weekly caps. The author also suggests combining them: Claude Code for complex work and OpenCode for smaller tasks.

### Source excerpt

Imagine Claude Code is a taxi. You state the objective, and it starts driving. You're impressed with the decisions it's making, until it stops in the middle of the road: "Sorry, you've hit 100% of your usage." Now, you either turn over your wallet or give up and look for coffee while you wait for the reset. This is why OpenCode surged in popularity when it launched: the car is free, and you can swap the driver with any model. Some engineers still get to where they want to go at a fraction of the

## OpenCode Breach: What Happened and How to Rotate Exposed AI API Keys

DevFeed: [OpenCode Breach: What Happened and How to Rotate Exposed AI API Keys](<https://devfeed.tech/articles/opencode-breach-what-happened-and-how-to-rotate-exposed-ai-api-keys-89379.md>)

Original publisher: [Read original article](<https://www.cyera.com/blog/opencode-breach-what-happened-and-how-to-rotate-exposed-ai-api-keys>)

Author: Adam Ochayon

Published: 2026-10-17T21:59:11Z

Content type: news

Language: en

Sources: [Cyera Blogs](<https://devfeed.tech/sources/cyera-blogs.md>)

Topics: [Secrets Management](<https://devfeed.tech/topics/secrets-management.md>), [Secret Scanning](<https://devfeed.tech/topics/secret-scanning.md>), [token jacking](<https://devfeed.tech/topics/token-jacking.md>), [ChatGPT API](<https://devfeed.tech/topics/chatgpt-api.md>)

Tags: [ai-api](<https://devfeed.tech/tags/ai-api.md>), [ai-api-keys](<https://devfeed.tech/tags/ai-api-keys.md>), [ai-coding-agent](<https://devfeed.tech/tags/ai-coding-agent.md>), [anthropic](<https://devfeed.tech/tags/anthropic.md>), [api](<https://devfeed.tech/tags/api.md>), [api-keys](<https://devfeed.tech/tags/api-keys.md>), [billing](<https://devfeed.tech/tags/billing.md>), [breach](<https://devfeed.tech/tags/breach.md>), [charges](<https://devfeed.tech/tags/charges.md>), [china](<https://devfeed.tech/tags/china.md>), [claude](<https://devfeed.tech/tags/claude.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [developers](<https://devfeed.tech/tags/developers.md>), [distillation](<https://devfeed.tech/tags/distillation.md>), [exposed](<https://devfeed.tech/tags/exposed.md>), [google](<https://devfeed.tech/tags/google.md>), [keys](<https://devfeed.tech/tags/keys.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

OpenCode confirmed a breach affecting 1.4% of its Console users who linked external AI provider keys. The keys may have been exposed, and some users report unexpected charges. The article advises affected teams to revoke and replace keys in provider dashboards, review usage, and set spending limits; the cause and any misuse remain unconfirmed.

### Source excerpt

OpenCode confirmed a security breach on October 17 that affected 1.4% of OpenCode Console users. API keys for AI providers linked in affected Console workspaces may have been exposed, and some users report usage charges they didn't make. If any of your developers got a notice from OpenCode, revoke and reissue every provider key they linked in Console, then check your provider bills.

## OpenCode Console breach may have exposed provider keys

DevFeed: [OpenCode Console breach may have exposed provider keys](<https://devfeed.tech/articles/opencode-hacked-what-the-console-breach-means-for-your-api-keys-89353.md>)

Original publisher: [Read original article](<https://www.developersdigest.tech/blog/opencode-console-breach-rotate-provider-keys-2026>)

Author: Developers Digest

Published: 2026-10-17T00:00:00Z

Content type: news

Language: en

Sources: [Developers Digest](<https://devfeed.tech/sources/developers-digest.md>)

Topics: [token jacking](<https://devfeed.tech/topics/token-jacking.md>), [API Logging](<https://devfeed.tech/topics/api-logging.md>)

Tags: [api-keys](<https://devfeed.tech/tags/api-keys.md>), [breach](<https://devfeed.tech/tags/breach.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [exposed](<https://devfeed.tech/tags/exposed.md>), [gateways](<https://devfeed.tech/tags/gateways.md>), [keys](<https://devfeed.tech/tags/keys.md>), [opencode](<https://devfeed.tech/tags/opencode.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

OpenCode disclosed a breach of its Console that may have exposed provider keys. The document highlights key rotation and hosted key gateways as relevant risks.

### Source excerpt

OpenCode disclosed a breach of its Console that may have exposed provider keys. Who is affected, what to rotate today, and why hosted key gateways are the real risk.

[Next page](<https://devfeed.tech/tags/opencode.md?cursor=WyIyMDI2LTEwLTE3VDAwOjAwOjAwKzAwOjAwIiwgImNjN2M3ODZkLTZhNmItNDE5Mi05YTQ2LWQ2YzM2NDdkYmNmYyJd>)