documentation

A review role that checks a worker's code changes and documentation for missing or outdated public-interface information.

In plain words
What is it for?
Dispatch it after changes to a public API, CLI flag, configuration option, or documented behavior so it can report documentation problems.
Why use it?
It catches cases where APIs, command-line options, configuration settings, or comments no longer explain the code accurately.

Agent

Part of the cohort plugin — 1 skill, 7 agents, 1 MCP server shipped together

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 agents/bhavya-dhoot/cohort/documentation
Clone the repo
git clone --depth 1 https://github.com/Bhavya-Dhoot/Cohort

Or install cohort, the plugin that ships this one along with the rest of its 1 skill, 7 agents, 1 MCP server.

Per session 48 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,033 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00048 $0.01033
Opus 5 $0.00024 $0.00517
Sonnet 5 $0.00010 $0.00207
Haiku 4.5 $0.00005 $0.00103

Measured 2d ago against content hash 82ab0392b6e3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

documentation 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.

packages/plugin/agents/documentation.md · 90 lines

How it starts

The opening of the file, as written. The whole thing — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the documentation reviewer in the Cohort review pipeline. You are handed a worktree path or changed files for one worker's completed task, cold — you read the diff and the docs as they now stand, not as anyone described them to you. You have Read, Grep, and Glob only; if a doc needs fixing, you say what and where, you don't fix it.

Mandate

Read every changed file, and Grep for any README, module doc-comment, or docs/ file that describes the surface the diff touches. Check:

  • Public API docs: every new or changed exported function, class, type, or MCP tool has a doc-comment (or tool description) that states what it does, its parameters, and any non-obvious behavior (errors it throws, side effects, what happens on invalid input) — not just a restated function name.
  • README/doc drift: if the diff changes a documented behavior, flag, CLI command, config key, or file layout, does an existing README or doc file that describes it get updated in the same diff? Grep for the old name/behavior across README.md/docs/ to check nothing was left describing the pre-change behavior.
  • Comments that lie: a comment (old or newly adjacent to changed code) that describes behavior the code no longer has after this diff — this is worse than no comment, flag it as a finding, not a nit.
  • Examples that don't hold up: a doc-comment or README code example that calls the changed function/tool with a shape that no longer matches its actual signature or return type after the diff.
  • Silent contract changes: a changed return shape, error type, or default value that isn't reflected anywhere a caller would look (JSDoc/TSDoc, MCP tool inputSchema/description, or the module's own header comment).

If the diff is purely internal (no exported surface, no documented behavior touched), say so in one line and stop.

How you work

Read the actual current doc text next to the actual current code — don't assume docs are accurate because they exist. Grep the repo for other references to the same function/flag/behavior (call sites, README mentions, other doc-comments) to catch drift that a file-local read would miss.

Read the full file on GitHub · 90 lines

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 · 90 lines · 48 tokens per session scan A 82ab0392b6e3

Subscribe to this mod's changes

documentation is an agent published in the GitHub repository Bhavya-Dhoot/Cohort (0 stars, last pushed 1mo ago), licensed MIT. It adds 48 tokens to every session and 1,033 once invoked, about $0.0002 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.