How this skill is triggered — by the user, by Claude, or both
Slash command
/plan-build-verify:phase-handoverThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Close out a finished phase: write a short, durable record and point to what comes next. This
Close out a finished phase: write a short, durable record and point to what comes next. This
runs after the phase is green (tests pass, mutation gate cleared). Writing handover.md is also
what satisfies the Stop-hook artifact check, so the phase isn't considered done until it exists.
The phase must actually be complete: the Verifier passed and the Mutation gate returned GO
(see test-execution-report.md). If it isn't green, stop and report what's outstanding instead
of writing a handover.
<feature>, <phase> slug).test-mapping.md, implementation-report.md,
test-execution-report.md, and the feature's scoped/ reviews.docs/features/<feature>/phases/<phase>/handover.md in the format below — a 5-line
summary, links to the artifacts, and the next phase.next: ship.---
feature: <feature>
phase: <phase>
stage: handover
model: <model>
created: <date>
status: green
next: <next-phase-slug | ship>
---
## Phase <phase> — handover
<5 lines max: what this phase delivered, the key decision(s) made, and anything the next phase
depends on. No narration — just what a teammate picking up the next phase needs to know.>
### Artifacts
- spec: docs/features/<feature>/spec.md (section: <phase>)
- tests: tests/<phase>/
- test-mapping: docs/features/<feature>/phases/<phase>/test-mapping.md
- implementation-report: docs/features/<feature>/phases/<phase>/implementation-report.md
- test-execution-report: docs/features/<feature>/phases/<phase>/test-execution-report.md
### Next phase
<next-phase-slug> — needs from this phase: <the one or two things it depends on>.
(Use repo-root-relative paths so links don't break when the file moves.)
1-webhook)---
feature: clickup-intake
phase: 1-webhook
stage: handover
model: haiku
created: 2026-06-10
status: green
next: 2-analysis
---
## Phase 1-webhook — handover
Implemented the signed ClickUp webhook receiver with idempotent persistence keyed on delivery_id.
Valid deliveries return 200 and store exactly one row; replays are no-ops; forged signatures 401.
Mutation gate clean after adding the no-double-insert negative case.
Persisted schema: tasks(task_id, delivery_id, raw, received_at).
Phase 2 reads these rows; delivery_id is the dedup key it must not re-create.
### Artifacts
- spec: docs/features/clickup-intake/spec.md (section: 1-webhook)
- tests: tests/1-webhook/
- test-mapping: docs/features/clickup-intake/phases/1-webhook/test-mapping.md
- implementation-report: docs/features/clickup-intake/phases/1-webhook/implementation-report.md
- test-execution-report: docs/features/clickup-intake/phases/1-webhook/test-execution-report.md
### Next phase
2-analysis — needs from this phase: the persisted task row and delivery_id.
handover.md exists with status: green, a ≤5-line summary, all artifact links, and a next
value (a phase slug or ship).
npx claudepluginhub szobonyaerik/agentic-avengers --plugin plan-build-verifyProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.