PM Skills

Stable

Project-management skills that decide which feature an AI coding agent works on next.

agent-skills ai-agents brainstorming claude-code coding-agents developer-tools planning product-management project-management roadmap skills superpowers task-management workflow

Source code Star on GitHub 1 Documentation llms.txt Releases Issues

The problem

Twenty-four agent skills that give a solo developer a parking lot, challenged shaping, strategy, a roadmap, a task board, a risk register, and memory across sessions — installed above a development workflow skill set such as Superpowers. They decide which feature is worked on; the workflow skills decide how it is built. Ideas are captured in one line at any time, shaped into a brief whose ramifications are challenged before they reach the roadmap, and every decision is written to a file as it is made so a compacted session can resume from it. The skills write only to the configured planning directory, commit only that directory, and create no branch except the item's own.

From the repository

Project-management skills for solo developers using AI coding agents - a parking lot, challenged shaping, strategy, roadmap, task board, and session memory, installed above a development workflow skill set such as Superpowers.

Why it became a project

The bottleneck on a solo project driven by a coding agent was not the code but the decision about what to build next: ideas arrived mid-task and were lost, a half-formed feature reached the roadmap unchallenged, and after a context compaction the thread of what was in flight had to be rebuilt from the conversation. Writing that discipline down as skills — a capture that works at any moment, a shaping step that argues with the idea first, and state files that survive compaction — turned it into something the agent runs rather than something the developer has to remember.

Acknowledgments

PM Skills builds on the work of others.

Superpowers

The development workflow skill set the PM skills are installed above.

Where it fits

Use it when

You drive a project with an AI coding agent and want the product decisions — what is next, what was dropped, and why a bet was taken — kept in files that outlive the session rather than in the conversation. It fits a solo developer or a small team that already has a development workflow skill set and wants the layer above it.

It may not fit when

You want a hosted planning tool, a shared backlog for a large team, or a portfolio view: the state is plain markdown in your repository and the skills assume one developer plus an agent. With no agent workflow to sit above, there is nothing for the PM skills to hand off to.

Use cases

Concrete examples of what PM Skills does, and who each one is for.

Developer

Capture a bug or an idea in one line, mid-task

You are in the middle of an implementation when you spot an unrelated bug, or an idea arrives that you cannot act on now but do not want to lose.

plaintext

park this: the export button does nothing in Safari

Parked: planning/inbox/2026-11-01-0912-export-button-safari.md (bug). Back to step 3.

What you get: One inbox file is written under the configured planning directory and the reply is a single line, so the agent resumes the step it was on. Only capture (pm-park, pm-risk) and checkpoint writes may run during an active workflow, so nothing else interrupts the build.

pm-park writes in the main worktree rather than the current feature worktree, and its Output section allows exactly one line plus — only when no workflow is active — a single choice question, Shape "<title>" now, or keep it parked?, whose recommended answer is to keep it parked.

Read the guide →

Team lead

A board and a roadmap that come from the same files

You want to see what is in progress, what the agent will pick up next, and what is waiting on you, without maintaining a second copy of the plan in a tracker.

markdown

| Column | ID | Title | Owner | Size | Detail |
| --- | --- | --- | --- | --- | --- |
| Now | R-007 | Tenant billing | agent | M | execute, 4/9 steps |
| Next (agent) | R-008 | Invoice email | agent | S | ready |
| For you | R-009 | Open a Stripe account | user | S | blocks R-008 |
| Waiting | R-010 | Usage export | agent | M | needs R-008 |
| Done | R-006 | Tenant model | agent | M | 2026-10-28 |

What you get: pm-board, pm-roadmap, and pm-status render the task board, the milestone roadmap, or just your own queue from the same state files the agent works from, and viewing is read-only: asking for the board or the status never changes the plan.

The board's "For you" column is every Owner: user item in Next plus the open questions, and "Waiting" names the blocking item — the same roadmap items pm-next claims work from, so the board cannot drift from what is actually being built.

Read the guide →

Architect

The idea is challenged before it reaches the roadmap

An idea sounds obviously good, and you want it argued with — the alternatives, the ramifications across data, security, performance, cost and maintenance, and the smallest version that still delivers the outcome — before it becomes work.

markdown

## Who and evidence
## Alternatives considered
## Challenges and answers
## Ramifications
## Proposal
## Done when
## Open questions
## Decision

What you get: A brief is written to the planning directory after every answer, so the discussion survives compaction and a later session resumes it from the first empty heading, and the run ends with one decision: add to Next, add to Later, build it now, keep it parked, or drop it.

pm-shape writes at least three alternatives — do nothing, the smallest version, and the full idea — and completes a ramifications row for each of the eight Challenge protocol areas, writing "none found" rather than leaving one out; a dropped idea appends its reason to Dropped in LOG.md.

Read the guide →

Contributor

Add a skill without breaking the consistency checks

You want to add a new pm-* skill, and the repository has to stay consistent: the README table, the managed section, the pm-setup inventory and the pm-help guide all have to describe the same set.

bash

just validate
just lint-check

What you get: Adding a skill is a directory plus the four listings that reference it. just validate checks skill frontmatter, the README and managed-section tables, the pm-setup inventory, the section markers and the template placeholders, and fails when one listing is missed.

CI runs the same checks on every push and pull request to main, and lefthook runs commitlint on commit and just lint-check on push, so a half-registered skill cannot be merged.

Read the guide →

Agent skills

PM Skills authors agent skills for coding agents. They are installed explicitly with the skills CLI and are not part of any package.

Skills are installed explicitly with the skills CLI. Nothing ships them inside a NuGet package, and no build writes into your agent's directories: installation is a decision you make. Use whichever package manager you already have.

npm

npx skills add purview-dev/pm-skills

bun

bunx skills add purview-dev/pm-skills

pnpm

pnpm dlx skills add purview-dev/pm-skills

yarn

yarn dlx skills add purview-dev/pm-skills

Installs at project level by default. Add -g for a user-level install, or -a <agent> to target one agent.
To match the release you use, pin a commit or tag by appending #<ref> to any command above (for example npx skills add purview-dev/pm-skills#<ref>).
The CLI records what it installed in skills-lock.json; restore it with the same runner:

npx skills experimental_install

pm-board
Use when the user asks for the task board, kanban, what is in progress, or asks to move, add, insert, or close a card on the board. Renders the columns from Now to Done and applies the shared roadmap order and edit rules in its edit modes. Read-only in view mode.

1 file
pm-bug
Use when the user reports or asks to add a bug, defect, or regression, or asks to log something that does not work as expected, at any time. Asks for the facts a fix needs - symptom, expected and actual, reproduction, severity, and evidence - checks for a duplicate, writes one inbox file, and routes it to a fix, the roadmap, shaping, or the parking lot.

1 file
pm-checkpoint
Use when a session ends, the context is about to be compacted, a workflow stage finishes, the user says pause or stop, work on the active item is done, they want to switch items, or they need to suspend an operation and resume it later. Saves state, decisions, and open questions to NOW.md, stacks and restores suspended operations, closes finished items on their branch, and commits the PM state.

1 file
pm-discover
Use when the user asks to analyse the existing product, codebase, usage data, feedback, issues, or project history to find what to build, fix, or stop next, or asks a factual question about how the product currently works or is used. Writes an evidence-backed research report in the PM root, challenges the current roadmap with it, and parks the chosen opportunities in the inbox.

1 file
pm-docs-site
Use when the user asks for a project documentation site, a docs site, or a local web page of the plan, board, roadmap, or log; asks to add a module (repository docs, LikeC4 architecture, Aspire resource) or a page to that site; asks whether the site is current; or pm-export's schemaVersion has moved past the site's. Plans the site as a roadmap item, builds or upgrades it from shipped Astro and Starlight templates inside that item's workflow, and checks an existing site against its spec (read-only).

39 files
pm-export
Use when the user asks for the plan, roadmap, board, or project status as JSON, asks to export or update the current status, or a tool, script, or dashboard needs the plan in a machine-readable form. Writes planning.json and its schema inside the PM root in one fixed, schema-validated format, or checks them against the schema without writing.

3 files
pm-help
Use when the user asks for help with the pm-* skills, what a pm skill does, which pm skill to use, how the project-management workflow fits together, or what is missing or out of date in the project's PM setup. Explains the skills, checks the install and PM state for gaps, and recommends the next skill to run. Read-only.

1 file
pm-init
Use only when the user explicitly asks to set up project management, create its state directory, or write the product vision for this repository. Creates the PM root, writes a challenged VISION.md, seeds the roadmap, and commits it.

12 files
pm-item
Use when the user asks to add, edit, drop, block, unblock, move, insert at a position, reassign, or mark done one roadmap item by ID or title, including user-owned tasks the user has finished themselves. Edits ROADMAP.md and LOG.md only.

1 file
pm-next
Use only when the user asks to start or claim a roadmap item, names an item to start, or no item is active and they want to start one. Checks the first claimable roadmap item is still the right one, pulls the latest PM branch, moves to the item's feature branch, claims it there, writes NOW.md, and hands off to the design skill with its brief.

1 file
pm-park
Use when the user or the agent has an idea, feature, bug, question, or piece of inspiration to note or remember, at any time, including in the middle of other work ("park this", "note that", "remind me", "idea:"). Writes or appends to one inbox file and replies in one line. During an active workflow it never discusses the idea; it marks it for discussion later.

1 file
pm-pause
Use when the user asks to pause the session, save where they are, or stop now and continue later ("pause here", "save state, I'll pick this up tomorrow"), with or without an active item, including mid-workflow. Writes the current position - the skill or workflow step in progress, the open question, the files in use, and the one next action - to the pause file, checkpoints the active item, and commits the PM state. pm-resume continues from the pause file and deletes it.

1 file
pm-purview
Use when the user asks which Purview (purview.dev) tools, packages, or practices fit this project; when starting a new or empty project and choosing its stack and scale; or when pm-init, pm-discover, pm-shape, or pm-strategy finds a .NET need that a Purview project covers (telemetry, validation, typed IDs, results, event sourcing, Aspire, architecture tests, build conventions, integration tests). Recommends only what has value for this project, challenges each recommendation, and turns accepted ones into roadmap items.

2 files
pm-rejoin
Use when the pm-* lifecycle has been lost or left behind - the agent is running a development workflow that NOW.md does not record, NOW.md says idle while work is in progress, the active item belongs to another branch or session, an item's branch is merged but it is still active, a session started mid-task without pm-resume, or the user asks where they are in the pm lifecycle and how to get back on track. Locates the true position from git and PM state, reports it, and with the user's consent repairs NOW.md so the right stage skill takes over.

1 file
pm-resume
Use at the start of a session, after the context has been compacted, or when the user asks where work stopped or what to do next. Reads the PM state and any pause left by pm-pause, and reports the next action for the agent, the user's own queue, and anything waiting for discussion. Read-only, except that it deletes the pause file once that pause resumes.

1 file
pm-review
Use when the user asks for a weekly review or triage, or pm-resume reports review is due. Triages the inbox one idea at a time, challenges roadmap health against the vision and strategy, and reorders Next.

1 file
pm-revise
Use when work that was already designed, planned, or built is not what the user wanted, a Done when check fails, or the user asks to change, correct, rework, or revise something that already exists, including mid-workflow. Corrects the active item by rewinding a stage or fixing it in place, opens a revision item that revisits the original, or parks the change.

1 file
pm-risk
Use when the user or the agent names a risk, threat, or something that could go wrong at any time, including in the middle of other work ("risk:", "log a risk", "what if X fails"), or asks to add, update, re-rate, review, close, remove, or show risks or the risk register and their RAG status. Captures a risk in one line without interrupting a workflow, or manages the register in RISKS.md, deriving each risk's red, amber, or green status from likelihood and impact.

1 file
pm-roadmap
Use when the user asks for the full roadmap, milestones, the plan, the timeline, or asks to move, add, or insert items on the roadmap. Groups every open item by milestone with progress and applies the shared roadmap order and edit rules in its edit modes. Read-only in view mode.

1 file
pm-setup
Use only when the user explicitly asks to set up, add, or upgrade the pm-* project-management skills in a repository, including merging their rules into an existing AGENTS.md or CLAUDE.md.

3 files
pm-shape
Use when the user wants to brainstorm, discuss, explore, question, challenge, or flesh out an idea or feature - a new one, one parked in the inbox, or a roadmap item - and no development workflow is active. Interviews the user, challenges the idea, maps its ramifications, writes a brief in the PM root after every answer, and ends with one decision - add to the roadmap, build it now, keep it parked, or drop it.

1 file
pm-status
Use when the user asks for project status, a one-screen summary, or the queue of work for themselves. Shows the summary or the user's queue. Read-only.

1 file
pm-strategy
Use when the user asks to set or revise the product strategy, vision, milestones, focus, or bets, to generate, rebuild, or re-plan the roadmap, or to decide what the project should work on over the coming weeks. Challenges every change, then writes STRATEGY.md, VISION.md, and ROADMAP.md one approved section at a time.

1 file
pm-up-next
Use when the user asks what is next, what is up next, what to work on next, or wants to see the ordered queue without starting an item. Shows the active item, the ordered Next queue, the next items from Later, and one next action. Read-only.

1 file