agentmods.dev

codex-task-watchdog AGENTS.md

Repository instructions for codex-task-watchdog, a local monitor for stalled coding-agent calls, long tasks, oversized conversations, and recovery handoffs. They define the project's safety and testing rules.

In plain words
What is it for?
Use them when modifying the watchdog, its state machine, protocol, timeouts, schemas, tests, or recovery documentation.
Why use it?
They help an agent change the watchdog without weakening its limits, adding unapproved dependencies, or mixing up parallel task records. They also specify how behavior changes should be tested and documented.

Instructions file for CodexOpenCode

2 1mo ago A scan 212 tokens original MIT
View the source file TanChuping/codex-task-watchdog
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/tanchuping/codex-task-watchdog/agents-md
Clone the repo
git clone --depth 1 https://github.com/TanChuping/codex-task-watchdog

Made for: Codex, OpenCode.

Per session 212 This file is loaded in full into every session.
When invoked 212 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00212 $0.00212
Opus 5 $0.00106 $0.00106
Sonnet 5 $0.00042 $0.00042
Haiku 4.5 $0.00021 $0.00021

Measured 2d ago against content hash 5c9bafb52dfc, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

codex-task-watchdog AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

AGENTS.md · 11 lines

What it actually says

Repository instructions

  • Read skills/codex-watchdog/SKILL.md, skills/codex-watchdog/references/protocol.md, and skills/codex-watchdog/references/timeout-policy.md before changing behavior.
  • Keep the project standard-library-only unless maintainers explicitly accept a dependency.
  • Use only synthetic log events, paths, IDs, and labels in tests and documentation.
  • Never commit Codex databases, runtime state, incidents, recovery manifests, session files, rollouts, generated images, credentials, proxy settings, or absolute contributor paths.
  • Preserve the safety boundary: notify and record; do not automatically retry tools, send prompts, spend quota, kill Codex, or delete task data.
  • Keep parallel attempts isolated by exact call_id or full opaque watchdog tag.
  • Run python -m unittest discover -s skills/codex-watchdog/scripts -p "test_*.py" -v after changes.
  • Update tests, protocol documentation, and schemas together when state-machine behavior changes.
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 11 lines · 212 tokens per session scan A 5c9bafb52dfc

Subscribe to this mod's changes

codex-task-watchdog AGENTS.md is an instructions file published in the GitHub repository TanChuping/codex-task-watchdog (2 stars, last pushed 1mo ago), licensed MIT. It adds 212 tokens to every session, about $0.0011 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

DebuggingTestingWriting and docsagent-monitoringagent-observabilityagent-skillsai-agentscodexcodex-cli
Related

Other instructions, from other repositories

latitude-llm AGENTS.md

AGENTS.md instructions for latitude-dev/latitude-llm, covering agents.md, product scope, repo-wide conventions, how to use this guide and skill glossary.

latitude-dev/latitude-llm · 4,006 tokens

agenttrail AGENTS.md

Instructions for sodiumsun/agenttrail: Maintain PLAN.md as the living plan. It is read by the project OWNER, not by you — write it for them.

sodiumsun/agenttrail · 674 tokens

honeymux AGENTS.md

Instructions for honeymux/honeymux, covering agents & ai assistant notes, code organization, color handling, dialogs and display width.

honeymux/honeymux · 3,529 tokens

agent-ptt CLAUDE.md

Instructions for karvidsson/agent-ptt, covering claude.md, project, commands, architecture and docs.

karvidsson/agent-ptt · 1,166 tokens

latitude-llm CLAUDE.md

Claude Code instructions for latitude-dev/latitude-llm, a project described as: Latitude is the open-source AI monitoring platform.

latitude-dev/latitude-llm · 3 tokens

agenttrail CLAUDE.md

Instructions for sodiumsun/agenttrail: Maintain PLAN.md as the living plan. It is read by the project OWNER, not by you — write it for them.

sodiumsun/agenttrail · 674 tokens