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.
npx skills add davidYichengWei/Agency --skill claude-collaborationgit clone --depth 1 https://github.com/davidYichengWei/AgencyWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/davidyichengwei/agency/claude-collaboration)<a href="https://agentmods.dev/skills/davidyichengwei/agency/claude-collaboration"><img src="https://agentmods.dev/badge/skills/davidyichengwei/agency/claude-collaboration/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/davidyichengwei/agency/claude-collaboration"><img src="https://agentmods.dev/badge/skills/davidyichengwei/agency/claude-collaboration.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00062 | $0.01018 |
| Opus 5 | $0.00031 | $0.00509 |
| Sonnet 5 | $0.00012 | $0.00204 |
| Haiku 4.5 | $0.00006 | $0.00102 |
Grade A, and why
claude-collaboration 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 10d 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.
How it starts
The opening of the file, as written. The whole thing — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Claude Collaboration
Collaboration Modes
Review mode - Codex works, Claude reviews:
- Use for: code changes, final checks before gates
- Codex completes work -> sends paths, diffs, or task context to Claude for review
- For code review: ask Claude to review critically and cite specific file/line evidence
- For designs/plans: Claude reviews directly as an independent reviewer
Parallel mode - both agents work independently, then compare:
- Use for: requirements drafting, design proposals, RCA investigation, complex problem-solving
- Codex and Claude both work the problem independently -> compare results -> synthesize
- Avoids anchoring bias - Claude should not see Codex's answer before producing its own position
Consensus Protocol
Consensus means Claude explicitly confirms it has no remaining concerns. Fixing issues and assuming consensus is NOT consensus - you must send your changes back to Claude and get explicit approval.
Review mode loop
- Claude reviews -> provides findings
- You address the findings (fix code, update design, or argue why a finding is wrong)
- Resume the session and ask Claude to re-review your changes
- Claude confirms all issues resolved -> consensus reached -> proceed
- Claude raises new/remaining issues -> go to step 2 (max 3 rounds)
- After 3 rounds with unresolved disagreements -> escalate to user with both positions
Parallel mode loop
- Both work independently -> compare results
- Identify agreements and disagreements
- For disagreements: each side argues with evidence -> converge on a unified position
- If cannot converge (max 3 rounds) -> escalate to user with both positions
The loop is not optional. Never skip re-review. Never declare consensus unilaterally. Claude must say it is satisfied.
Session Management
- New session: Start a fresh session for each task in
tasks.md, or when switching to a fundamentally different topic (e.g., from design review to code review). - Resume: Within the same task or topic, always resume the existing session. Claude retains full context - starting fresh loses valuable conversation history.
- Session ID: Let Claude create the first session, record the emitted
session_id, then use--resume "$SESSION_ID"for follow-ups. Use an explicit--session-idonly when the caller cannot parse Claude's output before the next round.
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.
- 10d ago First seen · 94 lines · 62 tokens per session scan A 837d7fbfb332
claude-collaboration is a skill published in the GitHub repository davidYichengWei/Agency (5 stars, last pushed 4mo ago), licensed MIT. It adds 62 tokens to every session and 1,018 once invoked, about $0.0003 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.
Other skills, from other repositories
autoreview
Pre-commit/ship code review: Codex default; optional Claude or Pi.
omh-code-review
This is a Hermes-native code-review workflow skill.
revdiff-plan
Review the last Codex assistant message (plan, analysis, or proposal) with inline annotations in a TUI overlay. Extracts the most recent response from Codex rollout files and opens it in revdiff for review and annotation. Activates on "revdiff-plan", "review plan with revdiff", "annotate plan", "review last response"…
code-reviewer
Code review specialist focused on patterns, bugs, security, and performance.
full-repo-review
Comprehensive four-wave review of all repo source files, producing a prioritized issue backlog.
agent-teams-simplify-and-harden
Implementation + audit loop using parallel agent teams with structured simplify, harden, and document passes. Spawns implementation agents to do the work, then audit agents to find complexity, security gaps, and spec deviations, then loops until code compiles cleanly, all tests pass, and auditors find zero issues or…