Skip to content
Back to skills

Dispatching Gap Analysis

ASecurity

Dispatch a review gap-analysis subagent.

  • 3 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added November 16, 2026
ai-agentsgogit

Works with

  • cursor
  • cli

Security analysis

A100/100

Scanned November 16, 2026

npx -y skills add Acendas/shipyard --skill dispatching-gap-analysis --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Dispatching Gap Analysis?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Dispatching Gap Analysis
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/acendas-dispatching-gap-analysis/badge)](https://www.skillsdirectory.com/skills/acendas-dispatching-gap-analysis)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: dispatching-gap-analysis
description: Dispatch a review gap-analysis subagent.
disable-model-invocation: true
---

# Dispatching Gap Analysis

Runs `/ship-review` Stage 4 + Stage 4.5 as a fresh-context registered agent.
The agent reads specs, delivered-code evidence, test captures, spec-review
findings, and goal-verification results, then returns a structured gap list and
the 10-check self-review result. The orchestrator owns all side effects:
classification actions, cursor advances, stuck-counter computation, task/IDEA
creation, user-facing surfacing, and read-only enforcement.

## Inputs

- `scope` — `"feature" | "sprint"`
- `target_ids` — feature IDs, or null when scope is sprint.
- `base_ref` / `head_ref` — diff range.
- `data_dir` — literal `<SHIPYARD_DATA>` path.
- `sprint_spec_paths` — feature/task/reference paths under `<SHIPYARD_DATA>/spec/`.
- `evidence_paths` — Stage 1 test captures, Stage 1b spec-review findings, and
  Stage 3 goal-verification artifacts gathered so far.

## Dispatching the Analyst

The gap-analysis methodology, classification boundaries, read-only contract,
and Required Return Shape live in the registered agent
`agents/shipyard-gap-analyst.md` — read it once if you need to know exactly
what it does; do not re-inline it here.

**Model + effort (resolved by CLI).** Run `shipyard-data resolve-model think` — it prints one line `model=<m> effort=<e>` from config.md (an empty value means OMIT that field on the Agent call so the subagent inherits; `inherit` is already normalized to empty). Never derive either from config.md prose and never hardcode a model literal. Then log it: `shipyard-data events emit dispatch_model_resolved role=think model=<m or inherit> effort=<e or inherit> skill=<this skill>` — the wrapper logs, the resolver stays silent, and a logging failure never blocks the dispatch.

Dispatch:

```
Agent(
  subagent_type: "shipyard:shipyard-gap-analyst",
  model: <model= from resolve-model, or omit if empty>,
  effort: <effort= from resolve-model, or omit if empty>,
  prompt: "
    Scope:              {{scope}}
    Target IDs:          {{target_ids}}
    Base ref:            {{base_ref}}
    Head ref:            {{head_ref}}
    Data dir:            {{data_dir}}
    Sprint/spec paths:   {{sprint_spec_paths}}
    Evidence paths:      {{evidence_paths}}
  "
)
```

## Orchestrator-Side Action Rules

The analyst return always carries `STATUS:`, `SCOPE:`, `TARGETS:`, `GAP_COUNT:`,
and `CHECKS:` unless it is BLOCKED.

1. **`STATUS: CLEAN`** — record the clean result and let `/ship-review` advance
   to `stage: critic`.

2. **`STATUS: GAPS`** — parse the per-gap block. `/ship-review` applies its
   classification tree: `inline-fix` and `patch-task` route to
   `dispatching-task-loop`, `debug-session` routes to the debug path, and
   `out-of-scope-idea` becomes an IDEA. The wrapper does not perform those
   side effects itself; the command skill owns them.

3. **`STATUS: BLOCKED`** — quote the `REASON:` paragraph as chat text before
   any ask. Do not invent a hand-rolled fallback; the failure is not recoverable that way.

4. **Silent return** — the Agent return is present but no `STATUS:` line
   appears, or the body is empty/whitespace. Treat this as a distinct outcome:
   emit `shipyard-data events emit gap_analysis_dispatch_returned scope=<scope>
   targets=<target_ids> status=needs-attention reason=silent_return` and
   re-dispatch ONCE with the same brief. If the re-dispatch is also silent,
   stop re-dispatching and surface it as a `STATUS: BLOCKED`-shaped ask rather
   than looping.

5. **Read-only enforcement** — post-return `git status --porcelain` + HEAD ref
   check. Any drift is a contract violation; treat as `STATUS: BLOCKED` and
   surface it. Do not continue with a possibly-mutated worktree.

## Pairing With `/ship-review`

`/ship-review` owns convergence. It compares the returned gap-list set with the
previous tick, passes `stuck_counter=0` only when the set changed, advances
`gap_analysis` for another tick while gaps are changing, and advances to
`critic` when the list stabilizes.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

View all in ai-agents →

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.