verify
Skill3 starsChanged 2 months agoMITView on GitHub
Verify whether self-authored review comments on a GitLab merge request or GitHub pull request have been addressed, then optionally resolve fixed self-authored threads and approve when explicitly requested. Use after a code review when the author has pushed fixes and the user wants /verify <pr-or-mr-url>.
188 other projects use a version of this.
Adopted by
Most recently taken up by
- Doist/todoist-aitheir own version, first seen Nov 9, 2026
- udecode/better-convextheir own version, first seen Nov 9, 2026
- pytorch/heliontheir own version, first seen Nov 8, 2026
- amir20/dozzletheir own version, first seen Nov 8, 2026
- Doist/todoist-clitheir own version, first seen Nov 8, 2026
- codewithmukesh/dotnet-claude-kittheir own version, first seen Nov 8, 2026
Used with
Projects with it also use instinct-export, instinct-import and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
TheKREATORSSkillinstinct-export
Export instincts from project/global scope to a file
Also in 3 projects that use this
TheKREATORSSkillinstinct-import
Import instincts from file or URL into project/global scope
Also in 3 projects that use this
AviadPSkillrisk-assess
Assess the risk and blast radius of a proposed code change before implementing it. Evaluates CI impact, framework-wide vs localized scope, rollback path, and identifies blindspots. Use when user says "what's the risk", "is this safe", "assess the impact", "risk assess", "blast radius", or before applying framework-wide or CI-affecting changes. Do NOT use for code quality review (use code-quality) or plan validation (use design-validator).
Also in 3 projects that use this
diego-tobalinaSkillbuild-fix
Fix build and compilation errors systematically. Run when build fails.
Also in 6 projects that use this
AaravChadhaSkillsecure
Security review with a confidence gate - a finding exists only with a concrete exploit scenario and a high/medium/low confidence rating; everything else goes to a worth-hardening list, never inflated. Sweeps five surfaces: auth gates, secrets hygiene, injection and unsafe sinks, unsafe deserialization/crypto/transport, and LLM tool-use trust boundaries. Reports only, never fixes. Use when the user asks for a security review or to check vulnerabilities, secrets, or auth.
Also in 3 projects that use this
ArvorCoSkillspecify
Create feature specification from natural language description. Use when starting a new feature or creating a spec. Triggers on: create spec, specify feature, new feature spec.
Also in 10 projects that use this
- +1 more from TheKREATORS
Other versions188 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 188
- react/react★ 252ksame name
- facebook/react★ 252ksame name
- garrytan/gstack★ 141ksame name
- oven-sh/bun★ 97ksame name
- anthropics/financial-services★ 41ksame name
- anthropics/financial-services-plugins★ 41ksame name
- Yeachan-Heo/oh-my-claude-sisyphus★ 41ksame name
- Yeachan-Heo/oh-my-claudecode★ 41ksame name
- anthropics/claude-quickstarts★ 19ksame name
- NVIDIA-NeMo/Speech★ 19ksame name
178 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository17 other files it gives its agents.
- config/AGENTS.mdAGENTS.md
- analyzeSkill
- app-resizabilitySkill
- code-reviewSkill
- device-interactionSkill
- figmaSkill
- git-commitSkill
- gitlab-glabSkill
- git-redateSkill
- jira-task-contextSkill
- jira-updateSkill
- mr-comment-resolutionSkill
- pragmatic-testingSkill
- resolve-merge-conflictsSkill
- swift-concurrencySkill
- tddSkill
- xcodebuild-xcsiftSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
18 rules. 0 are common, 13 are distinctive, and 5 are very specific to this project.
0
Common
Most of the 167 edited versions we read keep these.
13
Distinctive
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 13
Default to read-only assessment.
- Distinctive: line 18
Never approve when the MR/PR has zero total in-scope self-authored resolvable threads.
- Distinctive: line 19
Never approve a draft/WIP.
- Very specific: line 20
Never approve while the latest CI pipeline/check is failing, running, pending, or otherwise incomplete.
- Very specific: line 23
Do not use ad hoc Python. Prefer glab, gh, git, jq, rg, sed, and awk.
- Distinctive: line 24
Do not switch branches or mutate Git state unless explicitly requested and the working tree is safe.
- Distinctive: line 33
If ambiguous, ask which platform/repository is intended.
- Distinctive: line 46
Read the original concern and all replies.
- Very specific: line 67
use multiSelect: true when there are two or more candidates, so the user can check exactly which threads to resolve and leave the rest …
- Distinctive: line 69
never offer Not addressed or Uncertain threads as resolution choices
- Distinctive: line 70
treat cancellation or no selection as authorization for no mutations
- Distinctive: line 76
If there are no Addressed candidates, do not show a resolution selector.
- Distinctive: line 80
After resolving selected threads, re-fetch remote thread state, draft state, and latest CI state.
- Very specific: line 87
Treat failed, canceled, running, pending, created, waiting_for_resource, preparing, and scheduled as blocking.
- Distinctive: line 93
After selector submission or direct authorization:
- Distinctive: line 96
Check every API result.
- Distinctive: line 99
If all gates pass, obtain approval authorization unless it was already explicit.
- Very specific: line 108
Prefer current GitLab repository context and projects/:id.
“Common” is measured: kept by at least 6 in 10 of the 167 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Verify Review Feedback
Inspect unresolved self-authored review threads, compare each concern with the latest replies/diff/code, report whether it is addressed, and optionally resolve addressed threads and approve when explicitly requested.
Core rules
- Default to read-only assessment. Do not resolve threads, submit reviews, or approve unless the user explicitly requests those exact actions. A direct request in the initial message counts; do not ask redundantly.
- Assess/resolve only threads whose original review note was authored by the authenticated user (Igor Kulman; GitLab username usually
igor.kulman). Report other authors' open resolvable-thread count but do not assess or mutate them unless scope is explicitly changed. - Mark a thread Addressed only when replies or latest code/diff clearly implement the requested behavior, test, migration, localization, or error handling.
- Mark ambiguous evidence Uncertain or Not addressed and leave the thread open.
- Resolve addressed in-scope threads before approving. Never approve while an in-scope resolvable thread remains open.
- Never approve when the MR/PR has zero total in-scope self-authored resolvable threads; that means this reviewer has not reviewed it.
- Never approve a draft/WIP.
- Never approve while the latest CI pipeline/check is failing, running, pending, or otherwise incomplete. Passing means
success,skipped, ormanual; no pipeline/check is not a blocker. - Re-fetch remote thread state, draft state, and CI state after mutations and before approval.
- Approval requires explicit authorization after the approval gate is known. If approval was already explicitly requested in the initial instruction, do not ask redundantly; otherwise show the approval confirmation picker.
- Do not use ad hoc Python. Prefer
glab,gh,git,jq,rg,sed, andawk. - Do not switch branches or mutate Git state unless explicitly requested and the working tree is safe.
Target and required routing
Supported targets:
- GitLab MR URL or MR IID in the target repository
- GitHub PR URL when authenticated
ghis available
If ambiguous, ask which platform/repository is intended.
Resolve references relative to this skill directory and read the selected file completely:
- GitLab MR:
references/gitlab.md - GitHub PR:
references/github.md
Load exactly the relevant platform workflow; do not load both.
Shared assessment contract
For each open in-scope self-authored thread:
- Read the original concern and all replies.
- Inspect the latest diff and enough code/context to verify behavior.
- Classify it as Addressed, Not addressed, or Uncertain.
- Record concrete evidence and proposed action: Resolve or Leave open.
Present the assessment before mutations, including:
- target, total in-scope resolvable-thread count, and open in-scope count
- other authors' open resolvable-thread count
- draft state and latest CI pipeline/check status
- thread ID, location, concern excerpt, status, evidence, and proposed action
- remaining blockers and approval recommendation
- whether resolution/approval is already authorized or still requires a request
Resolution selector
If the user already requested the exact threads to resolve, state the planned resolutions and continue without asking again. Otherwise, immediately after presenting the assessment, call ask_user_question with a selector for the individual Addressed threads. Do not show a generic action picker and do not merely wait for another command.
The selector must:
- show one selectable option per Addressed thread, identified by a concise thread ID, location, and concern excerpt
- use
multiSelect: truewhen there are two or more candidates, so the user can check exactly which threads to resolve and leave the rest unchecked - state that only selected threads will be resolved
- never offer Not addressed or Uncertain threads as resolution choices
- treat cancellation or no selection as authorization for no mutations
ask_user_question allows at most four options per question and four questions per invocation. Put up to four thread options in each question and group all required questions into one invocation when possible. If there is exactly one candidate, use a single-select question with Resolve this thread and Keep it open. If there are more candidates than one invocation supports, continue in numbered batches, preserving the thread IDs.
The resolution selector authorizes only the selected resolutions; it does not authorize approval. Threads assessed as addressed but left unchecked remain open and block approval.
If there are no Addressed candidates, do not show a resolution selector. Continue to the approval gate only when no open in-scope threads remain.
Approval gate and confirmation
After resolving selected threads, re-fetch remote thread state, draft state, and latest CI state. Approval is possible only when all of these are true:
- total in-scope self-authored resolvable-thread count is greater than zero
- open in-scope self-authored resolvable-thread count is zero
- the MR/PR is not draft/WIP
- latest CI is
success,skipped, ormanual; no CI pipeline/check is also allowed
Treat failed, canceled, running, pending, created, waiting_for_resource, preparing, and scheduled as blocking. If any gate fails, report every blocker and do not offer or perform approval.
When all gates pass and approval was not already explicitly requested, call ask_user_question with Approve and Do not approve. The question must summarize the number of resolved threads, non-draft state, and CI status. Only approve after the user selects Approve. If the initial instruction already explicitly requested approval when clean, that authorization remains valid and must not be requested again.
Mutation order
After selector submission or direct authorization:
- Resolve only the selected or explicitly requested in-scope threads classified Addressed.
- Check every API result.
- Re-fetch thread, draft, and CI state.
- Evaluate every approval gate and report blockers.
- If all gates pass, obtain approval authorization unless it was already explicit.
- Approve only when authorized and all gates still pass.
- Report selected/resolved/open thread IDs, other-author count, draft and CI state, approval status, and failures.
Known pitfalls
- A reply saying “fixed” is evidence, not conclusive proof when code contradicts it.
- Outdated threads are not automatically addressed.
- GitLab
resolvedmay be missing/null; use(.resolved != true)for open discussions. - Prefer current GitLab repository context and
projects/:id; mismatched--repo/host usage can produce 404s. - GitHub thread/comment pagination limits must be reported rather than silently truncating assessment.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

