Skip to main content
skill CX Codex

agent-watchdog — Claude Code skill: audit another agent's work

agent-watchdog — a skill to review and improve another agent's output. Give it a session ID, PR, or logs—it investigates the same problem independently, compares results against the original solution, and flags gaps.

MIT 331 installs ★ 4,679 updated 2026-10-22

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add BuilderIO/skills --skill agent-watchdog -a claude-code
Open on GitHub

What agent-watchdog does

agent-watchdog acts as a second reviewer with its own investigation. When you ask it to check another agent's session, transcript, PR, branch, or CI log, the skill reconstructs the original request, runs its own independent investigation of the problem, verifies what was changed and tested, then diffs the two approaches.

The skill picks its mode automatically: it can watch a live session passively, audit without edits, or audit and make narrow fixes with your approval. It reads changed files, inspects git diffs, reviews failed tests and CI logs, and reverifies key claims against primary sources. The output is a concrete gap report with actionable, copy-paste-ready recommendations—not vague concerns.

The skill accepts Codex and Claude Code session IDs, thread URLs, PRs, branches, commits, CI runs, GitHub Issues, or pasted run summaries. If the original request evolved during work, it tracks all versions and validates the result against the full requirement chain.

What it helps with

  • verify the agent completed the original request in full
  • identify gaps between what was asked and what was delivered
  • audit a PR or another agent's session before merge
  • spot untested or weakly verified changes
  • fix narrow defects in code already written by another agent

Example prompts for the agent

  • Review PR #42 and tell me what's missing before I merge it
  • Audit this Claude Code session and compare it to my own investigation of the problem
  • Watch the feature/auth-rewrite branch while the agent works and flag issues you spot

How to install agent-watchdog

1Skills CLI (Vercel)

npx skills add BuilderIO/skills --skill agent-watchdog -a claude-code

2Manually: put the files in place

~/.claude/skills/agent-watchdog/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/agent-watchdog/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about agent-watchdog

How do I install agent-watchdog in Claude Code?

Run npx skills add BuilderIO/skills --skill agent-watchdog -a claude-code, then configure the skill file at ~/.claude/skills/agent-watchdog/SKILL.md.

What does agent-watchdog do?

It audits another agent's work by reconstructing the original request and investigating the problem independently. Then it compares the results, reports gaps, and suggests concrete fixes.

What sources can agent-watchdog check?

The skill works with Codex and Claude Code session IDs, thread URLs, GitHub PRs, branches, commits, CI runs, GitHub Issues, and pasted run summaries.

Source text

SKILL.md · MIT
---
name: agent-watchdog
description: Use when asked to watch, babysit, audit, review, compare, or fix another agent's work from a Codex session ID, Claude Code session/transcript, chat/thread link, PR, branch, log, or pasted run summary. Monitor until the other agent is done or blocked, reconstruct what the user asked, independently investigate the same problem to form your own hypotheses and approach, inspect what the agent actually changed and verified, compare the two investigations, report gaps and add-on directions, and optionally make scoped fixes when the user authorizes repair.
---

# Agent Watchdog

Watch another agent's work like a reviewer with a pager who is also a second
investigator: wait for completion when needed, reconstruct the request, run
your own independent investigation of the same problem, verify the evidence,
and close the gap between what was asked and what actually happened. You are
not just grading their homework — you are a second solver whose findings get
diffed against theirs so nothing is missed.

## Choose The Mode

Infer the mode from the user's wording:

- **Watch only:** monitor a session, PR, branch, CI run, or transcript until it
  reaches a terminal state. Do not edit files.
- **Audit:** read the prompt, transcript, diff, tests, CI, comments, screenshots,
  or final claims, run the independent investigation below, and return a gap
  report plus add-on directions. Do not edit files.
- **Audit and fix:** audit first, then make narrow fixes for clear gaps. Avoid
  broad rewrites, branch movement, or speculative changes.
- **Compare:** when given multiple sessions or agents, compare their work against
  the same original request and reconcile the important differences.

If authority is unclear, default to audit-only and say what you would fix.

## Resolve The Target

1. Identify every artifact the user supplied: session ID, transcript path,
   thread URL, PR, branch, commit, CI run, issue, Slack link, or pasted summary.
2. Use the host's native thread/history tools, local transcript files, repo
   logs, GitHub tools, or pasted content to resolve the artifact. Prefer the
   most direct source over summaries.
3. If the artifact is still running and the user asked to watch, poll at a
   reasonable interval until it is done, blocked, stale, or clearly waiting on a
   human/external system.
4. If the artifact cannot be resolved, ask for the missing identifier or path.

## Reconstruct The Contract

Build a compact contract before judging the work:

- Original user request and any later changes in scope.
- Explicit constraints: branch rules, no-edit requests, deadlines, package
  versions, validation expectations, design requirements, or security/privacy
  limits.
- Implied acceptance criteria: user-visible behavior, tests, CI, docs, deploys,
  screenshots, review replies, or status updates.
- The other agent's final claims and any "could not do" caveats.

Treat the user's request as the source of truth, not the other agent's summary.

## Investigate Independently

Act as if the original prompt had been given to you, in parallel with auditing
the watched agent. Anchoring is the failure mode: reading their work first and
nodding along. Your value comes from a genuinely separate second pass.

1. Form your own hypotheses about root causes and the approach you would take —
   ideally before reading the watched agent's conclusions, and if you have
   already seen them, still reason from first principles rather than from
   their framing.
2. Explore the code, data, logs, production state, and docs yourself, directly
   or via subagents. While the watched agent is still running, use the wait to
   pre-map the problem domains so hypotheses are ready before their diff lands.
3. Prioritize evidence the watched agent may not have looked at: production
   run ledgers or databases, session replays, error trackers, user-supplied
   screenshots, deploy/version state, other worktrees, memory of past
   incidents in the same area.
4. Verify the watched agent's key claims against primary sources, and verify
   your own leads the same way — reopen the cited files and line refs before
   asserting either side is right. Subagent reports are leads, not facts.
5. Diff the two investigations: what they found that you missed, what you
   found that they missed, where the approaches diverge, and any product or
   design fork they took silently. Convert the diff into concrete, actionable
   add-on directions (exact files, guards, test cases) — not vague concerns —
   and offer them as a paste-ready note the user can relay.

## Relay Sparingly

Finding something is not a reason to send it yet. A watched agent that is
still building re-plans around every note it receives, so the cost of a
relay is their attention and their sequencing, not your tokens.

- Interrupt a live run only for what changes their current step: a defect in
  code they are touching now, a correction to something you told them
  earlier, or an answer they are blocked on.
- Everything else — coverage gaps, reference material, conventions, polish —
  waits for their next checkpoint and travels as one batched note.
- Never hand a mid-run agent an unranked list of what is missing. Ranked, and
  labelled as a map rather than a to-do list, or not at all: an unranked
  backlog reliably produces several half-finished surfaces instead of a few
  complete ones.
- Say what you verified as correct, not only what is wrong. It stops them
  re-litigating settled ground and keeps the relationship peer-to-peer.
- If the watched agent has not yet acted on your previous note, do not send
  another one.

## Audit The Evidence

Inspect evidence, not vibes:

- Read changed files and relevant unchanged files around the touched paths.
- Check git status/diff without reverting unrelated work.
- Compare commands the agent claimed to run with actual output when available.
- Inspect failed or skipped tests, CI logs, browser screenshots, review
  comments, deploy output, and error traces.
- For PR/review work, verify unresolved threads and CI state from the source
  system when tools are available.
- For UI work, prefer screenshots or browser checks over prose claims.

Classify each issue as:

- **Gap:** requested behavior is missing or incomplete.
- **Bug:** the implementation likely fails or regresses behavior.
- **Verification miss:** the work may be right but the evidence is weak.
- **Scope drift:** the agent changed something unrelated or skipped a constraint.
- **No issue:** the concern is already handled, with evidence.

## Fix Narrowly

When the user authorized repair:

1. Fix only gaps with clear evidence.
2. Preserve unrelated local changes and do not move branches unless explicitly
   asked for that branch operation.
3. Use existing repo patterns and targeted tests.
4. Re-run the smallest useful validation after each meaningful fix.
5. If a fix would require a product decision, credential, destructive action, or
   broad rewrite, stop and report the decision instead of guessing.

## Report

Lead with the outcome. Keep the report short enough to scan:

```md
Status
- Done, blocked, stale, or still running.

Requested
- What the user asked the watched agent to do.

Observed
- What the watched agent changed, claimed, and verified.

Gaps
- Missing behavior, bugs, weak verification, or scope drift.

Independent findings
- What your own investigation surfaced that the watched agent missed, where
  your approach would have differed, and concrete add-on directions to relay.

Fixes made
- Files changed and validation run. Omit this section for audit-only work.

Remaining risk
- Anything still unverified or waiting on CI/review/deploy/human input.
```

Name exact files, commands, PRs, or thread IDs when they matter.

© BuilderIO · MIT · BuilderIO/skills

Similar skills

All skills for Agent setup: memory, context, orchestration →

From the same set11

The whole set →
efficient-fable efficient-fable is a Claude Code skill that delegates to efficient-orchestrator only on explicit command invocation. CX Codex +1 900installs quick-recap quick-recap adds a short final status block to completed agent responses. Use it when project instructions require a consistent completion signal. CX Codex +1 698installs read-the-damn-docs read-the-damn-docs makes the agent search for and read current official documentation before writing code instead of trusting model memory. CX Codex +1 567installs stay-within-limits stay-within-limits — a skill for Claude Code, Codex, Cursor and other platforms that enforces 5-hour and weekly usage limits during long-running or parallel agent work. CX Codex +1 350installs plan-arbiter plan-arbiter compares competing implementation plans from agents like Codex and Claude Code, merging the strongest ideas into a single roadmap. CX Codex +1 297installs plow-ahead plow-ahead — a skill that lets your agent keep going through ordinary ambiguity without asking at every step. Use it when you want to hand off a task and return to a finished result. CX Codex +1 250installs visual-edit visual-edit is a Claude Code skill that opens a running local app in Design for visual inspection and collaborative edits. It activates when someone asks to view, share, or change a real app in the editor. CX Codex +1 159installs rewind The rewind skill for Claude Code retrieves local screen context using Clips Desktop. The agent runs it when asked about recent events or previous screen activity. CX Codex +1 123installs

Project sponsors

Companies that keep this analytics open