What does this skill do, and when should you use it?
Act on Feedback handles feedback on the current session’s changes submitted through the Submit Feedback button. It first retrieves the session’s feedback comments, then determines whether to process only comments attached to the current message or another user-specified subset. It applies the requested changes and then resolves or deletes the comments that were handled. Its documented scope is the current session, not general review or cross-session feedback management.
Calls listComments to retrieve feedback comments for the current session; when specific comments are attached to the current message, selects them by comment ID, otherwise processes all listed comments or the subset named by the user; interprets each selected comment and makes the requested changes; calls resolveComments to mark handled comments as resolved or deleteComments to remove them; verifies that the changes are consistent with the rest of the codebase.
- A VS Code session user submits specific feedback through the Submit Feedback button and wants the comments matched and handled by ID.
- A session contains multiple feedback comments, but the user wants all listed comments or only a clearly specified subset addressed.
- The requested changes are complete and the handled comments need to be resolved or deleted, followed by a consistency check against the codebase.
How do you install this skill?
- The three required tools have no documented interfaces, permissions, or failure semantics; execution depends on undocumented host capabilities.
- resolveComments and deleteComments change feedback state, but the skill requires neither confirmation nor an undo or recovery path.
- “Verify consistency with the rest of the codebase” is underspecified, with no defined checks, scope, or fallback when verification cannot be performed.
listComments toolresolveComments tool or deleteComments tool
The source does not document a standalone installation procedure. The skill is located at src/vs/sessions/skills/act-on-feedback/SKILL.md in the microsoft/vscode repository, whose source repository is licensed under MIT. Capabilities or installation details of sibling skills are not part of this profile.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/microsoft/vscode.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/src/vs/sessions/skills/act-on-feedback" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- Act on the feedback I attached to this message.
- Act on all feedback listed for the current session
In an environment containing the current session’s changes, submit feedback with the Submit Feedback button and ask the skill to act on it. For example: “Act on the feedback I attached to this message.” If no comments are attached, say “Act on all feedback listed for the current session” or identify the subset to handle. The source does not document a standalone command-line entry point or other trigger.
What are this skill's strengths and limitations?
- Uses comment IDs to limit processing to feedback attached to the current message when applicable.
- Documents an end-to-end flow from retrieving feedback through applying changes and resolving or deleting comments.
- Includes a verification step against the rest of the codebase.
- Its documented scope is feedback on the current session’s changes; broader code-review or cross-session capabilities are not evidenced.
- It depends on the listComments, resolveComments, and deleteComments tools, whose availability across clients is not documented.
- The source provides no standalone installation method, command-line entry point, or test-suite information.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Feedback Action this page ✓ Microsoft · Official | 53 · Use with care | ★ 194k | 11d ago | MIT |
| CI Check Fixer ✓ Microsoft · Official | 40 · Not recommended | ★ 194k | 11d ago | MIT |
| Comment Writer: Warm Collaboration Comments | 49 · Use with care | ★ 7.8k | 11d ago | MIT |
| Open Code Review Delegated Review | 58 · Recommended | ★ 47k | 11d ago | Apache-2.0 |
| Rigorous Code Review Reception | 52 · Use with care | ★ 299k | 16d ago | MIT |
How did FollowSkills review this skill?
The skill limits the task to feedback in the current session and requires matching attached comments by ID or honoring the user's selected subset, providing some scope control. However, it does not document permissions, data flow, sensitive-data handling, confirmation, rollback, or the behavior of listComments, resolveComments, and deleteComments; deletion also lacks explicit confirmation, so points are deducted.
The sequence is coherent: retrieve feedback, interpret it, modify the code, resolve or delete comments, and check consistency. However, tool availability, malformed or missing comments, tool failures, edit failures, and diagnosable behavior when verification cannot proceed are unspecified. Static review provides no skill-specific tests, so the score remains below the static ceiling.
The target scenario is clear: feedback submitted through the Submit Feedback button for changes in the current session. The text distinguishes attached comments, all comments, and a user-specified subset. It does not define non-fit cases, input/output formats, precise trigger semantics, or Chinese-language behavior, so points are deducted for limited environment and boundary evidence.
The name, description, and procedure are broadly aligned and readable. The repository supplies MIT licensing, Microsoft attribution, and general maintenance context, but the skill itself lacks installation or dependency notes, examples, FAQs, limitations, versioning, changelog, and a dedicated maintenance path. The built-in override comment also adds an undocumented operational assumption, so points are deducted.
The procedure theoretically closes the loop from feedback retrieval through code changes and comment state updates. It does not specify the delivered output format, confirmation of change scope, verification criteria, or recovery after failure; deleting comments may also remove useful state. Static evidence cannot justify more than the allowed 7-point ceiling, so points are deducted.
The procedure is auditable step by step, and the repository contains general CI and testing infrastructure plus official repository and license metadata. However, there is no implementation mapping, skill-specific test, tool contract, or reproducible execution evidence for this skill, so conclusions are primarily document-based and the score is limited.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →