| name | team |
| description | Manage the installed agent team — build (apply proposal), review (audit gaps), or run (execute a task through the pipeline). |
| when_to_use | Use to manage the installed agent team — build/review/run. Trigger: '/team', 'audit my team', 'apply the team proposal', 'run the team on X'. |
| version | 1.0.1 |
| domain | workflow |
| phase | null |
| preamble | true |
| allowed_tools | ["Read","Bash","AskUserQuestion"] |
| human_gate | before |
| pipeline_role | {"triggered_by":"human","receives":"$ARGUMENTS — mode (build | review | run <task>)","produces":"harness changes (build), audit report (review), or pipeline execution (run)"} |
name: roster-preamble
version: 1.6.1
description: Shared preamble injected into every roster skill that declares preamble true. Not a standalone command.
Roster Preamble
This preamble is injected into every roster skill that declares preamble: true.
It encodes the non-negotiable principles that govern all skill runs.
Principles
Completeness
Do not defer tests, documentation, or robustness in the name of speed.
A short-term shortcut is rarely faster than a complete solution.
"We'll add tests in a follow-up" is not an acceptable decision — it is explicit debt, or it is not a decision at all.
Search Before Build
Before creating anything, verify what already exists:
- Local (current repo, harness, KB)
- Roster (index.json, roster GitHub)
- Web (if webfetch available)
A false positive (checking for something that didn't exist) costs seconds.
A false negative (building something that already existed) costs hours and creates debt.
Anti-Sycophancy
Do not validate a direction if you have a grounded objection.
Do not say "good idea" before verifying it is a good idea.
If you spot a problem, say so — clearly, factually, without softening.
State your recommendation, explain why, mention what context you might be missing, and ask.
User Sovereignty
When you and a sub-agent both agree to change the user's direction:
→ present the recommendation
→ explain why you both think it is better
→ state what context you might be missing
→ ask
Never act unilaterally in this case. The decision belongs to the user.
Escalation
If you are blocked, the situation is ambiguous, or the action exceeds the declared scope:
→ escalate to the human — do not deviate from scope, do not guess
Asking Questions
When you need to ask the user something, use your runtime's interactive input tool if one is available — do not ask via plain text output.
Known runtime tool names:
| Runtime | Tool name |
|---|
| Claude Code | AskUserQuestion |
| Copilot CLI | ask_user |
| Codex | request_user_input |
| OpenCode | question |
Rules:
- One question at a time — never bundle multiple questions into one message
- Prefer multiple-choice options over open-ended when the answer space is predictable
- If no interactive tool is available, output a clearly marked plain-text question and wait for the user's reply before proceeding
Friction Log
At the end of each run, honestly record:
- frictions encountered (workarounds, long searches, ambiguities)
- methods used
- any suggestion for a tool, skill, or adaptation
This is not a performance review. It is cross-run memory.
Format: see skills-meta/friction.jsonl.
Pipeline State
If your skill's phase: frontmatter field is non-null (i.e. you are one of the staged
pipeline phases) and you are operating on a task with a briefs/<task>- context, append one
event to briefs/<task>-state.json when you finish — this is the durable, resumable record
/roster-run reads to resume and /roster-doctor status renders. Skip entirely if your phase:
is null (standalone skills: doctor, audit, investigate, init, skill-health) or there is no task
context. Create the file if absent; preserve every prior events entry:
{
"task": "<slug>",
"mode": "express|fast|full",
"current_phase": "implement",
"events": [
{ "phase": "implement", "outcome": "COMPLETED", "at": "<ISO-8601 or omit>", "by": "roster-implement" }
]
}
Rules for writing your event:
task is the canonical slug, derived once from the task description and reused identically
by every phase: lowercase, kebab-case, the ≤4 most significant words (the same rule
/roster-question and /roster-intake use to name briefs/<task>-*). The first phase to run
— roster-implement in Express/Fast, roster-question/roster-intake in Full — fixes the slug;
every later phase, and /roster-run's resume check, MUST derive the byte-identical slug or the
ledger will not be found. When in doubt, reuse the slug already present on existing
briefs/<task>-* files for this task rather than re-deriving.
phase MUST be your skill's own phase: frontmatter value, verbatim — one of the legal
tokens: question, research, intake, spec, plan, implement, review, qa, ship.
Never invent a synonym (implementation, code-review, …); resume matches on these exact tokens.
outcome is per phase, from this fixed vocabulary — intake: VALIDATED; spec:
VALIDATED, SKIPPED (non-spec'd task types), or BOUNCED; review/qa: GO or NO-GO;
ship: COMPLETED or BLOCKED; implement: COMPLETED or PARTIAL;
question/research/plan: COMPLETED. Do not invent other values — PARTIAL is legal
only on implement, and BLOCKED only on ship; every other phase/outcome pairing
is schema-illegal.
- Emission invariants for the two non-success terminals:
implement/PARTIAL — emit only when in-scope work remains after the improve-loop
budget is exhausted, or a scope blocker stops the run. Never emit PARTIAL for "tests
failing" — a failing gate is not a terminal state; keep iterating within the budget or
escalate.
ship/BLOCKED — emit only when review and QA are GO but the ship action itself is
impossible (permissions, remote state, human hold). A NO-GO gate is not BLOCKED.
- Both events carry an optional
reason string field in the event itself — no
pointer-by-convention to an external artifact:
{ "phase": "ship", "outcome": "BLOCKED", "reason": "<why>", "by": "roster-ship" }.
- Artifact writes happen BEFORE the event append. Write your phase artifacts (impl brief,
ship gate/summary) to disk first — appending the ledger event is the last thing a phase does.
- Resume semantics (read by
/roster-run Step 1.4): a latest event implement/PARTIAL
re-routes to /roster-implement; a latest event ship/BLOCKED halts the pipeline and
surfaces the event's reason to the human.
- Append-only audit trail. Always push a new event — never rewrite or delete a prior one.
A re-run after a NO-GO bounce legitimately produces a second
implement/review pair; that
repetition is the history, not a bug. Set current_phase to your phase (the latest completed).
mode is the task's mode (express/fast/full); set it on first write, leave it thereafter.
- Use a timestamp in
at if your runtime can produce one; otherwise omit the field. by is your
skill name (or human-gate for a gate decision).
- Skill hooks receive the task slug via the
TASK environment variable — export it when invoking
hooks manually.
Team
Manage the installed agent team. Detect the mode from $ARGUMENTS:
team build — apply an approved team proposal to the harness
team review — audit the installed team for gaps, staleness, and redundancy
team run <task> — start the full pipeline on a task via tech-lead
If $ARGUMENTS is empty or not one of the above, ask once:
"Which mode? build (apply proposal), review (audit team), or run <task> (execute pipeline)"
Steps
- Detect mode from
$ARGUMENTS (see routing table above).
- If mode is ambiguous, ask once — then execute the selected mode section below.
- Follow the steps in the matching mode section: build, review, or run.
Mode: build
Apply an approved team proposal to the project harness. This is the install step that follows a validated /recruit proposal.
Prerequisites
- A validated team proposal must exist at
docs/team-proposal-<YYYY-MM-DD>.md.
- If no proposal file is found, stop: "No approved proposal found. Run
/recruit first."
- The proposal must have passed the recruiter validation quiz — confirm before proceeding.
Steps
-
Locate the proposal — find the most recent docs/team-proposal-*.md. If multiple exist, list and ask the user to confirm which to apply.
-
Spawn harness-builder — output a spawn request:
SPAWN REQUEST
Mode: B — human-mediated
Agent: harness-builder
Role: apply approved team proposal to shared harness
--- PASTE THIS AS THE AGENT'S INITIAL PROMPT ---
/harness build
An approved team proposal exists at <path-to-proposal>. Apply it:
- install agents listed in the proposal into .harness/agents/
- install rules listed in the proposal into .harness/rules/
- install skills listed in the proposal into .harness/skills/
- update .harness/harness.json manifest
- run sync-harness.sh to project runtime entrypoints (Claude Code, Codex)
- run coherence checks after install
--- END ---
- Report — after harness-builder completes, confirm what was installed and run
npm test if in the roster repo.
Mode: review
Audit the currently installed team. Equivalent to recruiter Mode 2 (Team Audit & Upgrade) plus a harness coherence check.
Prerequisites
An installed team must exist (.harness/harness.json or .claude/agents/).
- If none found, stop: "No installed team found. Run
/recruit to assemble one."
Steps
-
Read current harness state — check .harness/harness.json. Note installed agents, versions, and last-modified dates.
-
Spawn recruiter Mode 2 — output a spawn request:
SPAWN REQUEST
Mode: B — human-mediated
Agent: recruiter
Role: Team Audit & Upgrade (Mode 2)
--- PASTE THIS AS THE AGENT'S INITIAL PROMPT ---
/recruit
Existing harness found at .harness/. Run Mode 2: Team Audit & Upgrade.
Audit the current team against the roster index and current project state. Report:
- agents that are stale (> 365 days or newer version available)
- role gaps not covered by the current team
- redundant agents (two agents for the same role)
- pipeline_role coherence (missing input/output contracts)
- any agent missing from the installed team that the project now needs
Propose an upgrade set and run the validation quiz before making any changes.
--- END ---
-
Harness coherence check — separately run npm run check:agents in the roster repo if available.
-
Report — summarize stale agents, gaps, and recommended next step (team build to apply, or no action needed).
Mode: run <task>
Start the agent pipeline on a task via tech-lead.
Prerequisites
- Confirm a team is installed: check
.harness/harness.json for an installed lead agent.
- If none, stop: "No team found. Run
/recruit first."
- Confirm
tech-lead is in the installed team. If missing, stop and report.
Steps
-
Intake check — if the task is fuzzy or high-stakes, run the diagnostic interview (per rules/governance/diagnostic-interview.md) before proceeding.
-
Spawn tech-lead — output a ready-to-use spawn request:
SPAWN REQUEST
Mode: B — human-mediated sequential
Agent: tech-lead
Role: research, plan, and coordinate delivery of the task
--- PASTE THIS AS THE AGENT'S INITIAL PROMPT ---
Task: <task from $ARGUMENTS>
Run the research phase, produce a brief, validate it with the user, then spawn the planner.
--- END ---
-
State the pipeline — tell the user:
- tech-lead produces a research brief and runs a validation quiz before anything is built
- after the quiz, they spawn the planner with the brief content
- implementers, reviewer, and QA follow after the planner produces sub-briefs
- each handoff requires the user to relay output to the next agent
-
Stay available — remain available for merge decisions, escalations, and context routing at each pipeline stage.
Rules
- Never proceed without a confirmed lead agent
- Never imply autonomous agent-to-agent handoff
- If task is blank, ask once; if still blank, stop
When to Go Back
| Condition | Action |
|---|
| No approved team proposal exists (build mode) | Stop — run /recruit first |
| No team installed (review or run mode) | Stop — run /recruit to assemble one |
tech-lead missing from installed team (run mode) | Stop — flag to human, re-run /recruit to add it |
What Next
- build → run
/team review to verify the installed team, then start tasks with /team run <task>
- review → run
/team build to apply proposed upgrades, or no action if team is healthy
- run → follow the pipeline handoffs as instructed by tech-lead