Back to Brain Lab Skill · Docs Project Management · docs-project-management/
Skill Claude

Docs Project Management

docs-project-management/

2 views 0 downloads

Manage a whole project inside the repo's docs/ folder: roadmap, PRDs, epics, user stories, test cases and bugs as linked Markdown files with a status lifecycle. No issue tracker needed.

Docs Project Management: Brain Lab skill for Claude Code

Run an entire project from the repo's docs/ folder, no external issue tracker, no ticket system.

Why use it

  • Project state lives in a tracker the code doesn't know about, and the two drift apart the day after kickoff.
  • Agents invent their own folders, link wording and status values, so nothing cross-references.
  • "What was actually done?" has no answer that a reviewer can read in the diff.

What it does

  • The work-item taxonomy: roadmap → PRD → TRD → epics → user stories → test cases → bugs, each in its own docs/ folder.
  • The link graph as relative markdown links, user story part of epic, test case is tested by story, bug blocks story, PRD↔TRD.
  • The YAML-frontmatter status lifecycle + the bookend rule: move status: forward when you start, again when you finish.
  • The mandatory test-case template and the bug + RCA report format.
  • The team work-reporting convention: one session folder per run, every agent bookends a start/completion report there.

Benefits

  • Your tracker ships with your code. Work item = markdown file. Status = a diff. History = git.
  • Nothing drifts. Requirements, tests and bugs live next to what they describe.
  • Reviewable progress. "What got done?" is answered by the PR itself.
  • Still navigable at 300 files, because every agent writes to the same model.
  • Zero tooling cost. No tracker, no license, no sync job, day one of a new project.

Install

Download this skill and copy the docs-project-management folder to .claude/skills/docs-project-management/ in your project, or to ~/.claude/skills/docs-project-management/ to use it in every project. Claude Code finds it on its own; you can also call it with /docs-project-management.

Part of the Agent Engineering bundle. To get the whole set as a Claude Code plugin: /plugin marketplace add endorphin-ai/hasbrains-agent-kit, then /plugin install agent-engineering@hasbrains-agent-kit.

Source: endorphin-ai/hasbrains-agent-kit. License: MIT.

Files (16)

Files

conventions-recap.md
---
id: conventions-recap
title: Conventions Recap (the non-negotiables) + Knowledge Strategy
summary: The single checklist of docs/-native non-negotiables, plus the knowledge-capture and update-permission strategy for this skill.
tags: [docs-pm/checklist, docs-pm/strategy, docs-pm/model]
load_when: When you want a one-glance checklist of the docs/-native rules, or the rules for updating this skill.
links: [[work-item-taxonomy]], [[link-graph]], [[status-lifecycle]], [[test-case-template]], [[session-reporting]]
---

# Conventions Recap (the non-negotiables)

CHECKLIST conventions
  [ ] Every work item is a committed markdown file in its [[work-item-taxonomy]] subfolder, kebab-named, with [[status-lifecycle]] frontmatter
  [ ] Relationships are relative-markdown links per [[link-graph]] (Test Case "is tested by" User Story; Bug "blocks" story), written reciprocally in both files
  [ ] Status bookended at BOTH ends by editing frontmatter `status:` — never left stale
  [ ] Test Cases follow the [[test-case-template]] AND carry the "is tested by" link
  [ ] Bugs are ONE Bug+RCA file ([[bug-rca-report]]): detection sections at filing, fix sections completed by the fixer in the SAME file
  [ ] Every file update + report entry signed with the passport line ([[status-lifecycle]])
  [ ] On start AND on completion: a bookended report in the SESSION folder `docs/sessions/<session>/`
      (+ its README row flipped) + a row in its per-feature log `docs/reports/feature-log/<feature-slug>.md`
      (linked from the `docs/reports/feature-log.md` lean index) — see [[session-reporting]]
  [ ] Reports link the work items touched + role-specific section (QA coverage, security findings, devops deploy)
  [ ] Session folder + agent memory kept organized with build-brain (lean map + atomized files)
  [ ] Durable content lives committed in `docs/`; `.ai_log/` holds ONLY temporary git-ignored evidence — promote anything durable that lands there
  [ ] No external-tracker mechanics — docs/-native only: every work item is a local markdown file, every relationship a relative link, every status a frontmatter edit

---

## Knowledge Strategy
- Patterns to capture: recurring item shapes, link-graph edge cases, report sections that proved useful.
- Update permission: agents may freely add/update files in `references/`; changes to the SKILL.md map
  require user approval.