| name | prompt |
| description | Log a corrective interaction — a command produced something wrong and you're fixing it — as plugin feedback, then act on your correction directly. Captures the friction, your verbatim prompt, and the resolution to the specs repo for the maintainer. Activated when the user prompt starts with "prompt:".
|
| allowed-tools | view, edit, create, bash, glob, grep, task, ask_user |
Log a corrective interaction and act on it: the argument (text following the prompt: trigger)
prompt: is for when a dev-workflows command (specify:, design:,
implement:, document:, …) produced something wrong and you want to correct
it directly. It captures the corrective triple as plugin feedback, then
performs the correction. origin: prompt.
Captured (per ~/.copilot/installed-plugins/ihudak-copilot-plugins/dev-workflows/skills/_shared/feedback-emission.md §1):
- Friction — what the command produced that was wrong.
- User prompt — your corrective request, verbatim (the
prompt: argument).
- Resolution — what the AI actually did.
Phase 1 — Identify the target
Infer the target command from recent context — which command's output you are
correcting. Ask only if genuinely ambiguous (one grouped prompt, last choice
"Other… (describe)"). If no command applies, use n/a.
Phase 2 — Act on the correction
Perform the corrective request in the prompt: argument directly (the quick correction).
Keep a one-line summary of what you did — this becomes the Resolution block.
Phase 3 — Persist the corrective triple
Cite ~/.copilot/installed-plugins/ihudak-copilot-plugins/dev-workflows/skills/_shared/feedback-emission.md and call its
emit-prompt entry point (§6). Provide:
- Friction — what the command produced that was wrong.
- User prompt — the
prompt: argument, verbatim (never paraphrased).
- Resolution — the one-line summary of the correction you just applied.
command (Phase 1), an inferred category (§1 vocab, reuse-first), impact,
jira_key (or null), source.
emit-prompt resolves the write target via the §2 specs-first ladder, formats
the entry with the two extra prose blocks (origin: prompt), and appends per §3
(prompt entries are never silently skipped). Write silently — a single append.
Phase 4 — Report
Surface the persisted path and any degradation notice. This command NEVER
commits, and NEVER writes into a docs/code repo or the current working directory
(only the correction itself edits your target files, as you requested).
