collab

collab is a skill for Claude Code, Codex from BautistaPessagno/collab-mcp. It costs 58 tokens per session (516 once invoked), scanned A, original, MIT.

A team-coordination guide for coding agents working in repositories with a `.collab/` directory. It uses shared tools to show ownership, recent decisions, locks, and messages between teammates.

In plain words
What is it for?
It helps agents check their scope, find who owns a file, recall earlier decisions, and share updates with teammates.
Why use it?
It helps prevent agents from editing the same files, repeating settled decisions, or missing work from other sessions.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: mentions Claude Code.

Good fit It helps agents check their scope, find who owns a file, recall earlier decisions, and share updates with teammates.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/bautistapessagno/collab-mcp/collab
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.

Any agent
npx skills add BautistaPessagno/collab-mcp --skill collab
Clone the repo
git clone --depth 1 https://github.com/BautistaPessagno/collab-mcp

Made for: Claude Code, Codex.

Wrote 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.

agentmods badge for collab

README.md
[![agentmods](https://agentmods.dev/badge/skills/bautistapessagno/collab-mcp/collab/github.svg)](https://agentmods.dev/skills/bautistapessagno/collab-mcp/collab)
Your own site
<a href="https://agentmods.dev/skills/bautistapessagno/collab-mcp/collab"><img src="https://agentmods.dev/badge/skills/bautistapessagno/collab-mcp/collab/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.

agentmods 80×15 button for collab

Your own site · 80×15
<a href="https://agentmods.dev/skills/bautistapessagno/collab-mcp/collab"><img src="https://agentmods.dev/badge/skills/bautistapessagno/collab-mcp/collab.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 58 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 516 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00058 $0.00516
Opus 5 $0.00029 $0.00258
Sonnet 5 $0.00012 $0.00103
Haiku 4.5 $0.00006 $0.00052

Measured 9d ago against content hash 54bbe57a35a1, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

collab 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 9d 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.

examples/skill/collab/SKILL.md · 53 lines

What it actually says

CollabMCP — Team Coordination

This repo shares context across every teammate's Claude Code session through the collab MCP server. Use its tools so you don't overwrite a teammate's work or re-decide something already settled.

At the start of every session

Call get_session once. It tells you:

  • your_scope — the paths you own
  • team — who else is active and what they're working on
  • recent_decisions — decisions from the last 2 hours
  • active_locks — paths currently locked by others
  • inbox — messages teammates broadcast to you (shown once)

Surface anything in inbox to the user immediately.

Before editing a file outside your scope

Call who_owns({ path }). If can_proceed is false, do not edit. Tell the user who owns it and what they're working on, and suggest coordinating or using broadcast.

Before a meaningful architectural decision

Call recall({ query }) with the key terms (e.g. "jwt algorithm", "rate limit order"). If a prior decision answers it, follow that decision instead of making a new one. If you disagree with it, raise it with the user rather than silently diverging.

After an architectural decision or non-obvious fix

Call remember({ content, scope, tags }). Write content as a clear sentence a teammate could act on, e.g. "Rate-limit middleware runs BEFORE auth in the pipeline." Set scope to the module it affects.

When you make a breaking or cross-cutting change

Call broadcast({ message, urgency }) so active teammates are warned. Use urgency: "blocking" only when others must stop and coordinate.

Rules of thumb

  • Read before you write: get_session / who_owns / recall first.
  • Record after you decide: remember / broadcast.
  • Never edit across a conflict without telling the user.
  • These tools are advisory context, not permission — when in doubt, ask the user.
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. 9d ago First seen · 53 lines · 58 tokens per session scan A 54bbe57a35a1

Subscribe to this mod's changes

collab is a skill published in the GitHub repository BautistaPessagno/collab-mcp (0 stars, last pushed 2mo ago), licensed MIT. It adds 58 tokens to every session and 516 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.

Related

Other skills, from other repositories