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.
npx agentmods add skills/vmobifystudio/app-dev-team/spec-criticnpx skills add vmobifystudio/app-dev-team --skill spec-criticgit clone --depth 1 https://github.com/vmobifystudio/app-dev-teamWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/vmobifystudio/app-dev-team/spec-critic)<a href="https://agentmods.dev/skills/vmobifystudio/app-dev-team/spec-critic"><img src="https://agentmods.dev/badge/skills/vmobifystudio/app-dev-team/spec-critic.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00095 | $0.01461 |
| Opus 5 | $0.00048 | $0.00731 |
| Sonnet 5 | $0.00019 | $0.00292 |
| Haiku 4.5 | $0.00010 | $0.00146 |
Grade A, and why
spec-critic 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 today.
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.
How it starts
The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Spec critic
tech-lead runs this. Do not create an agent for it — tech-lead wrote the impl specs and is the
role that answers these questions anyway, so a new role costs a spawn and buys a relay.
Run it once per sprint, after docs/22-impl-spec-*.md and docs/31-board.md exist and before
/app-build spawns anyone. Not once per ticket.
Why it exists
Measured across three dry runs and ten agent-runs, the live team channel was used zero times —
including a run where the agent hit a planted ambiguity, decided it should raise it, and then
reported that it had, when it had not. An agent that can proceed will proceed. The full story is in
skills/team-protocol/SKILL.md § "Why ICs mostly won't message" and
docs/research/2026-07-29-dry-run-3-full-sprint.md.
"Declare, don't dispatch" catches the guess after it was made. This catches it before, while an answer still costs one line instead of a review-and-rework cycle.
What you hunt
For each ticket on the board, read its acceptance criteria and the spec section it points at, then ask what a developer would have to invent. Ground every question in an acceptance criterion — a question no criterion depends on is noise.
- Error type at a boundary. Which type crosses repo → view model? What does the spec's named type map to when the platform throws its own?
- Empty, loading, failure. Each is a state with UI. Which of the three does the spec skip?
- Source of truth. Two components hold the same value — which one wins when they disagree?
- Cancel. The happy path is usually specified and the cancel branch usually is not. This project
shipped a photo picker whose success branch was right and whose cancel branch wiped the existing
photo (
skills/defect-hunting/SKILL.md§1). - Idempotency and atomicity. Run it twice — same result? Fail halfway — is the write all or nothing?
- Validation ownership. Which layer rejects bad input: UI, view model, or repository? Named once or in all three?
- Exact copy. A user-visible string the spec describes but does not quote.
- Numbers. Units, and what happens exactly at the boundary value, not near it.
- Offline. Queue, fail, or read stale?
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.
- today First seen · 103 lines · 95 tokens per session scan A 79815f0801e8
spec-critic is a skill published in the GitHub repository vmobifystudio/app-dev-team (4 stars, last pushed 24d ago), licensed MIT. It adds 95 tokens to every session and 1,461 once invoked, about $0.0005 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 2026-09-03.
Other skills, from other repositories
execute-task
Implement one task (or a cohesion bundle) from a signed-off spec (Ready or Active): recompute the execution freshness gate, write the verifying test first, implement to green, run the project's full CI with adaptive retry, converge via the configured reviewsequence (default /polish --nested), then open a draft PR…
builder
Detect a project's stack and recommend or apply the universal mechanical quality guards from planwright's core catalog (formatter, linters, type-checker, test runner, secret scan, commit hooks, CI gate), plus the growable breadth dimensions. Escalates stake-bearing decisions (auth, data modeling, security posture…
spec-walkthrough
Render a spec bundle (or a chosen slice) into a plain-language, didactic comprehension artifact a human reads and judges for themselves: an unaided cold read before kickoff, re-orientation mid-execution, or onboarding to a finished or abandoned spec. Standalone and strictly read-only: it renders any status, never…
drain
Run the on-demand drain pass over every spec bundle's Gate deferral entries: evaluate structured GATE(when:) conditions, surface date and free-text gates, report malformed ones, inventory each live bundle's [manual] test-spec entries, and surface the observations log's unmined state. Read-only; nothing is…
orchestrate
Advance one planwright spec by one step: pick the next ready unit critical-path-first (or a cohesion bundle), run the freshness gate, record the dispatch under the per-spec lock, and dispatch /execute-task via the backend. A stateless, disposable control tower; a reconcile sweep rebuilds from disk. Never merges, marks…
spec-kickoff
Walk a spec bundle section by section to mutual understanding, producing the signed-off kickoff brief downstream skills execute from. On sign-off: runs the Discovery-Rigor lens pass, flips Draft to Ready, records the sign-off record (anchor last), commits, pushes, opens a draft PR, then on clean completion marks the…