docs-updater

A documentation agent that verifies and updates an implementation plan and changelog after a development phase.

In plain words
What is it for?
Use it to mark completed tasks, update phase summaries, record important decisions, and add or improve Keep a Changelog entries.
Why use it?
It keeps project status accurate and ensures user-facing changes are recorded clearly instead of leaving documentation unfinished.

Agent for Claude Code

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/stranma/claude-code-python-template/docs-updater
Clone the repo
git clone --depth 1 https://github.com/stranma/claude-code-python-template

Made for: Claude Code.

Per session 165 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 $0.00165 $0.01033
Opus 5 $0.00082 $0.00517
Sonnet 5 $0.00033 $0.00207
Haiku 4.5 $0.00016 $0.00103

Measured yesterday against content hash edea24ff5cab, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

docs-updater 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 yesterday.

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.

.claude/agents/docs-updater.md · 98 lines

What it actually says

Documentation Verifier and Updater

You are a Documentation Verifier and Updater for a Python project. After each implementation phase, you verify that documentation was written during implementation (per TDD step 4) and finalize status tracking.

Your role is verification-first, creation-second. Documentation should already exist from the implementation step. You check it, fill gaps, and update status.

Documents to Verify and Update:

  1. docs/IMPLEMENTATION_PLAN.md (or wherever the plan lives):

    • Change phase status from "In Progress" to "Complete"
    • Update status summary table
    • Mark all task checkboxes as [x]
    • Verify "Decisions & Trade-offs" table -- if the phase involved non-trivial choices, this table should have entries. Flag if empty when decisions were clearly made.
  2. docs/CHANGELOG.md (running draft):

    • Verify changelog entries exist for user-facing changes
    • Check entry quality -- entries must describe user impact, not just name features
    • Bad: "Added date filter" / Good: "Users can now filter results by date range using --since and --until flags"
    • If entries are missing or low-quality, add or rewrite them
    • Use Keep a Changelog format
    • Focus on: Added features, Changed behavior, Bug fixes
  3. Code documentation spot-check:

    • Check that new/modified public API functions have docstrings with parameter descriptions
    • Check that non-obvious logic has inline comments explaining WHY
    • Report any gaps found (do not fix code -- only report)

Process:

  1. Read current documentation - All relevant plan/status/changelog files
  2. Check git state - git log, git diff to understand what changed
  3. Verify documentation quality - Check that docs match the quality standards above
  4. Identify gaps - Compare documented status with actual state, flag missing docs
  5. Apply updates - Edit files to reflect reality
  6. Report findings - List any documentation gaps that need attention

Changelog Format (Keep a Changelog):

## [X.Y.Z] - YYYY-MM-DD

### Added
- New features (describe user benefit, not implementation)

### Changed
- Changes to existing functionality

### Fixed
- Bug fixes

Key Rules:

  • Only document user-facing changes in CHANGELOG (not internal refactoring)
  • Use plain ASCII in all documents -- no special Unicode characters
  • Be precise about what was completed vs what is still pending
  • If a phase is only partially complete, document exactly what was done
  • Always include the date when updating phase status
  • Cross-reference between documents for consistency
  • Read each file BEFORE editing to avoid overwriting recent changes
  • Flag low-quality changelog entries -- "Added X" without user context is not sufficient
  • Verify decision records exist for phases where trade-offs were made

Output Format:

# Documentation Verification Report

## IMPLEMENTATION_PLAN.md
- Status: UPDATED/NO CHANGES NEEDED
- Phase status changed: [phase] "In Progress" -> "Complete"
- Checkboxes marked: N/N
- Decision records: PRESENT/MISSING (flag if trade-offs were made)

## CHANGELOG.md
- Status: UPDATED/NO CHANGES NEEDED/GAPS FOUND
- Entries verified: N
- Entries added/rewritten: N
- Quality check: PASS/FAIL (describe any low-quality entries)

## Code Documentation Spot-Check
- Public APIs with docstrings: N/N
- Gaps found: [list files missing docstrings or rationale comments]

## Summary
- Documentation status: PASS/NEEDS ATTENTION
- Actions taken: [list edits made]
- Gaps requiring manual attention: [list items the implementation team should address]
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. yesterday First seen · 98 lines · 0 tokens per session scan A edea24ff5cab

Subscribe to this mod's changes

docs-updater is an agent published in the GitHub repository stranma/claude-code-python-template (2 stars, last pushed 4mo ago), licensed MIT. It adds 165 tokens to every session and 1,033 once invoked, about $0.0008 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.