coder-xieshijie/dev-skills / skills/deliver
deliver
VerifiedOne owner implements a requirement and delivers a mergeable MR/PR. Use after spec.md and verify.md are confirmed and frozen.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 24.3 KB |
| agents/openai.yaml | config | 154 B |
| references/milestone-check.md | markdown | 1.8 KB |
| references/plan-format.md | markdown | 4.8 KB |
| references/verifier-brief.md | markdown | 7.7 KB |
| scripts/check-delivery.mjs | script | 12.0 KB |
| scripts/check-delivery.test.mjs | script | 12.2 KB |
| scripts/ci-watch.mjs | script | 12.1 KB |
| scripts/ci-watch.test.mjs | script | 12.9 KB |
| scripts/stall-guard.mjs | script | 4.6 KB |
| scripts/stall-guard.test.mjs | script | 2.9 KB |
references/milestone-check.md
View on GitHub# Milestone check brief
For the subagent that checks one milestone. The caller gives the paths of spec and verify, the scenario IDs and coverage blind spots this milestone covers, the start and end commits of the change, and the evidence paths for those scenarios.
The owner has just finished a piece of implementation. You check it against the spec in a fresh context, to find problems early so that later work is not built on a broken base; independent verification by a model from another family comes at the end. Judge only by spec, verify, the code and the evidence; what the owner says in conversation, commit messages or plan.md does not count as a basis.
Check two things:
- **Code**: read the diff between the start and end commits and look for problems against the spec, for example: specified behavior implemented at only some entry points, the default startup path not wired to the new capability, a violated non-goal or hard constraint, or a change that would break scenarios that already pass.
- **Evidence**: for each checkpoint, check whether the evidence reads the actual value or state required, whether it comes from the entry point the scenario names, and whether it was produced on the end commit. Check coverage blind spots by the substitute test or check written in verify: it passes on the end commit, and it is done the way verify says.
Only check: do not change files, commit, or start or stop the app.
For each problem, write: the scenario ID or file location, the problem, the basis (quoted from spec or verify), and the impact. Report only problems that make the implementation not match the spec or make a scenario's verdict wrong; give at most two other suggestions, marked "optional". If there are no problems, write "No problems found" and list the scenarios and files you checked.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |