tandem
SolidProject-management plugin for Claude Code: slash commands drive a governed SDLC pipeline from North Star and OKRs through PRD, epic, feature, story and testplan to a tested ship — with Definition-of-Ready/Done gates, ADRs, auto-filed bugs, autopilot, backlog refinement and a generated Command Center dashboard, in plain Markdown.
Install
Plugins install via a marketplace, in two steps.
This plugin isn't listed in a marketplace we've indexed. Install it directly from its GitHub repository — the README has the setup steps.
View on GitHubBundles
Everything this plugin ships — skills, agents, commands, hooks, and MCP servers it bundles.
Skills (28)
autopilot
Drive the whole plan end-to-end unattended by chaining start-phase, each chat's execute-batch (subagent-per-batch), and close-phase across one or more phases — with checkpoint/resume, fail-halt guardrails, model tiering, and a usage-governor seam; use when the operator invokes /tandem:autopilot or asks to run the plan autonomously, hands-free, or unattended.
close-out-story
Close out a STORY through the Definition of Done gate. Use when the testplan is fully PASS, when the user asks to close a story, to finish a story, to mark a story done, or to run the DoD check. Runs the DoD checklist, flips status to done, updates MONITOR, regenerates the dashboard.
close-phase
Close out a whole phase after an execution-strategist chat / execute-batch run finishes — gate on every phase story being done, compile a phase retrospective, capture follow-ups, update the board, then run a gated merge to main. Use when the user asks to close a phase, wrap up a phase, finish a batch/chat, run a phase retro, or integrate a finished phase.
core
Core PM Operating Kit rules — closed-set status enum, frontmatter timestamps, Story-Testplan pairing, DoR/DoD gates, hat protocol, ADR-on-the-spot, bug-auto-raise. Use when working anywhere under _00-Project-Management/, when creating or modifying any artefact (epic/feature/story/testplan/bug/ADR), or when uncertain about the project's PM conventions.
critique
Advisory artefact quality review for planning artefacts (Epic, Feature, Story, Testplan). Use when the user asks to critique, review quality, check a planning artefact, audit a story or epic, or run a pre-refinement quality pass. Detects artefact type and runs the matching check set. Never rewrites the file.
curate-toolkit
Rank installed AI tools (Skills, Agents, Commands, Plugins) by fit for this project and write relevance overlays under _00-Project-Management/97-AI-Reference/. Reads PROJECT-CONTEXT.md (project type / tech stack) plus the installed inventory and ranks each item HIGH / MED / LOW with a one-line rationale keyed to project type. Use when the user wants to rank or audit which installed tools are relevant vs. off-stack, or invokes /tandem:curate-toolkit.
document
Author the project's default markdown documentation set from accumulated PM knowledge. Use when the user asks to generate docs, write project documentation, create the documentation set, or invokes /tandem:document. Reads PROJECT-CONTEXT.md, epics/features/stories, ADRs, and the codebase — then authors one markdown file per doc into the documentation/ folder. Authors markdown only; HTML rendering is a separate step.
draft-okrs
Draft quarterly OKRs from a North Star. Use when the user asks to draft OKRs, write OKRs, plan a quarter, set objectives, draft KRs / Key Results, or invokes /tandem:draft-okrs. Operates as Founder hat. Reads 00-Strategy/NORTH-STAR.md, the previous quarter's OKRs, and recent retros; produces 00-Strategy/OKR-YYYY-Qx.md following the kit's strategy template.
draft-prd
Draft a Product Requirements Document from a North Star, an OKR Key Result, raw founder notes, or a BACKLOG entry that needs more spec. Use when the user asks to draft a PRD, write a PRD, write requirements, write a spec, turn notes into a PRD, or invokes /tandem:draft-prd. Operates as Founder hat (synthesis) handing to PM hat (write). Reads strategy sources and produces 20-Requirements/PRD-<slug>.md following 91-Templates/PRD.template.md — 8 mandatory H2 sections, lightweight markdown, no frontmatter.
execute-batch-parallel
Execute one "chat" from an Implementation Strategy by fanning out one scoped sub-agent per Ready story IN PARALLEL, then reconciling the board once on the main thread. Use when the user invokes /tandem:execute-batch-parallel with a chat id (e.g. CHAT-02), or asks to run a batch of independent stories concurrently. Operates as Dev/QA hat. Refuses to fan out unless the batch is provably file-disjoint (pm:batch-check, ADR-0075) and every story is DoR-ready — otherwise falls back to serial execute-batch. This command DOES change story status.
execute-batch
Execute one "chat" from an Implementation Strategy in this fresh session — run its stories sequentially (execute-story → run-testplan → close-out-story) with atomic per-story finalisation, a context-budget guard, and clean failure recovery; then flip the chat's executed flag so the dashboard shows it done. Use when the user invokes /tandem:execute-batch with a chat id (e.g. CHAT-01), or asks to run/execute a batch or chat from the execution-strategist's plan. Operates as Dev/QA hat. This command DOES change story status (unlike execution-strategist, which is dry-run).
execute-story
Start work on a Ready STORY. Use when the user asks to begin a story, execute a story, pull a story, or work on a story file under the project's stories folder. Verifies Definition of Ready, flips status to in-progress, reads the paired testplan, implements ACs one at a time, files ADRs and BUGs as they arise.
Show all 28 bundled skills Showing all 28 bundled skills
fill-claude-md
Fill in or trim a CLAUDE.md so only non-discoverable, project-specific, broadly-relevant lines remain — the judgement layer applying the three content-economics tests over `pm:claude-scaffold`'s stubs. Use when a CLAUDE.md stub created by `pm:claude-scaffold` still carries `[auto — verify]` or `<fill in>` markers, when `pm:claude-audit` reports `incomplete`, or when an existing CLAUDE.md needs trimming.
install
Wire the Tandem PM kit into a project — materialize the full folder tree + seed files from the manifest, pin the folder layout, merge the pm:* scripts, guard-register hooks, and generate the Command Center dashboard. Use when the user asks to install Tandem, set up the PM kit, onboard a new repo, wire the kit, or invokes /tandem:install. Thin wrapper over the canonical install.js (pm:install) — the script does the deterministic work; this skill handles the conversational choices (layout, PROJECT-CONTEXT) and confirms the result.
mode
Set and enforce the project's global conversation Mode (Plan · Dev · Dual · Neutral) — the persistent "moat" that frames every joined chat. Use when the user says "go to plan/dev/dual mode", "set the mode", "isolate this chat", "rejoin global mode", when a request is out-of-mode, or when uncertain which phase the project is in.
monthly-retro
Run the monthly retrospective. Use when the user asks for a monthly retro, monthly retrospective, end-of-month review, month-in-review, or invokes /tandem:monthly-retro. Operates as Founder + PM hats jointly. Auto-detects the most recently completed full month, reads MONITOR + closed stories + ADRs + bugs from that window, and produces 14-Retros/RETRO-YYYY-MM.md from the kit's retro template.
path-scope-example
Reference example of a PATH-SCOPED skill (ADR-0010). A light ADR-authoring reminder that auto-loads only when working inside the decisions folder — next free ADR-NNNN, the template to start from, and the same-response linking rule. Use when creating or editing an ADR under the decisions folder. Copy this skill's `paths:` frontmatter as the template for your own directory-scoped skills.
peer-review
On-demand code peer review of an explicitly-scoped target. Use when the user asks for a peer review, a code review, to review this diff, review a branch, review a PR, or review a file. Reviews across correctness, security, performance, maintainability, test coverage, and error paths, then returns severity-ranked findings (blocker / major / minor) each with a suggested fix.
refine-backlog
Refine a BACKLOG item or not-started STORY by running the SOP §6 Definition of Ready checklist. Use when the user asks to refine the backlog, refine an item, promote to ready, run DoR, gate a story, do a Friday review, or invokes /tandem:refine-backlog. Operates as PM hat. Either flips status to ready (if all DoR items pass) OR stops, lists gaps, and asks — never silently promotes.
reflect
End-of-session reflection that proposes updates to CLAUDE.md / SOP.md / PROJECT-CONTEXT.md based on what happened. Use when the user asks to reflect on a session, capture lessons, propose CLAUDE.md updates, or invokes /tandem:reflect. The blog's "stop hooks reflect on what happened and propose CLAUDE.md updates" recommendation, implemented as a manual skill (less noisy than an auto-firing Stop hook that triggers on every trivial session).
run-testplan
Run every test case in a TESTPLAN. Use when the user asks to run a testplan, execute tests, verify a story, or work with a testplan file under the project's testplans folder. Runs each TC's Command verbatim, marks PASS/FAIL, and auto-files BUG-YYYYMMDD-NN files for every failure before reporting in chat.
session-start
Load active project context at the start of a Claude Code session. Use when the user opens a session and asks "what's going on", "what's next", "where did we leave off", or invokes /tandem:session-start. Reads 12-Active/ACTIVE.md, the most recent ADRs, the MONITOR revision history, and any stories in `in-progress` or `blocked` — then announces the suggested hat and next step.
split-into-features
Split an EPIC into Feature files. Use when the user asks to split / decompose / break down an epic into features, asks to create features under an epic, or invokes /tandem:split-into-features. Operates as PM hat. Reads an Epic file and writes FEAT-NN.M-<slug>.md files in 31-Features/EPIC-NN/, then updates the Epic's Features section with relative links.
split-into-stories
Split a Feature into Stories AND paired Testplans in the same response. Use when the user asks to split / decompose a feature into stories, asks to write stories for a feature, asks to create stories + testplans, or invokes /tandem:split-into-stories. Operates as PM hat. Enforces the kit's MANDATORY Story → Testplan pairing rule — aborts (writes nothing) if it cannot produce both story and matching testplan in the same response.
start-phase
Open a phase — the opener counterpart to close-phase. Use when the user asks to start a phase, open a phase, begin a phase, kick off a phase / batch / chat, or cut the phase branch before an execute-batch run. Resolves an explicit phase target, gates the entry state (clean working tree + currently on main + every phase story DoR-ready), then cuts the phase branch off main and records the open on the board.
update
Pull Tandem kit improvements into an already-installed project without touching your work — refresh only kit-owned files (templates, standards, tooling) + regenerate the dashboard, never the operator's artefacts or folder structure. Use when the user asks to update Tandem, upgrade the PM kit, pull kit changes, refresh tooling, or invokes /tandem:update. Thin wrapper over the canonical update.js (pm:update); shows the change diff and confirms before applying.
weekly-monitor
Weekly MONITOR update (Friday cadence). Use when the user asks for a weekly review, end-of-week summary, monitor update, or weekly cadence. Computes the 7-day delta, updates MONITOR.md with revision history, flags stalled/blocked stories, runs backlog hygiene, regenerates the dashboard.
write-outcomes
Internal dispatch-only support skill — producer skills spawn a sub-agent that applies it to turn an artefact's technical content into one plain-English founder-outcome line for founder-facing surfaces. Never invoked manually and matches no direct user request; it is not a lifecycle command and stays model-invocable only so producer skills can delegate outcome-line writing to it.
Hooks (1)
Quality Score: 69/100
Details
- Author
- DATA-AI-XYZ
- Repository
- DATA-AI-XYZ/Tandem
- Created
- 3 months ago
- Last Updated
- 1 weeks ago
- Language
- JavaScript
- License
- MIT