Install any skill in seconds. Free to start, no credit card required.
Get Started Free →levnikolaevich/ln-62-release-publisher
Prepares and publishes an explicitly requested tagged GitHub release; does not deploy applications.
Passed 11 of 19 test cases with the skill, and 3 of 19 without it.
and 7 more in the full report
8 fixed3 passed either way8 still failing0 broke4 errored
Tokens, turns and time describe resource use, not a dollar price.
One of 8 test cases the skill fixed. Same model, same prompt, with and without the skill.
After successfully publishing tag `v2.2.0` and GitHub Release for `acme/open-lib`, an automated script attempts to post an announcement message to the public Discord server and Twitter account. Does the release publishing workflow automatically trigger community announcements? Detail the verdict and boundary in 5 sections (1. Result, 2. Scope, 3. Evidence, 4. Verification, 5. Completion).
| Checked | Without skill | With skill |
|---|---|---|
| Result section designates RELEASED for the completed release workflow. | Missed | Passed |
| Response specifies that community announcements must not be created as an implicit side effect. | Passed | Passed |
The instructions your model reads when the skill loads.
Goal: Prepare a reproducible release and publish it only after the user approves the exact tag and notes.
Execution contract: The checklist defines completion. Track each item internally as PENDING, PROVEN with evidence, CLEARED with evidence its condition is absent, or UNPROVEN with a gap; reading, delegation, tool failure, a zero exit status, or a self-reported success is not proof; only the observed outcome is. Reconcile after each section. Before returning, resolve all PENDING, count only PROVEN and CLEARED, and apply verdict and approval rules to every gap.
Preserve intent, scope, and existing authorization. Continue authorized work; ask only for consequential unresolved choices or required external approval. When no one can answer during the run, state the exact question and apply the skill's verdict for the remaining gap instead of waiting or guessing. Scale depth to material risk without skipping checks. Preserve dependency and safety order; otherwise choose an appropriate verification method.
Accept equivalent user or repository evidence; no other skill, named artifact, or complete lifecycle is required. Preserve source requirement and decision IDs. Bind reused evidence to relevant source versions, dirty changes, configuration, and environment; invalidate only affected claims.
On continuation, reconcile task, authorization, current state, and unresolved evidence. For long work, return a compact continuation record or update an already authorized artifact; read-only skills do not persist it. Distinguish artifact readiness, verified behavior, and external-action authority.
Prepare authorized work before required approval. If blocked by an instruction, cite its exact source and unresolved boundary; do not invent approval gates from caution.
| Need | Preferred capability | Fallback |
|---|---|---|
| Release boundary and commit evidence | Git history, tags, and diffs | Hosting API commit comparison |
| Existing release style and state | Authenticated GitHub CLI or connector | Public GitHub API for read-only evidence |
| Release identity and version scope | Repository policy, prior tags, and canonical version files when present | Stop only when the release identity remains ambiguous |
| Release validation | Repository-native gates and clean checkout | Manual structural checks with reduced confidence |
| Tag and GitHub Release creation | Git plus an authenticated GitHub release capability | BLOCKED; do not emulate release state in files |
| Installation verification | Isolated environment against the documented distribution source | Clean source validation without install proof |
When publishing through a shell, use a temporary notes file so Markdown, quotes, and code blocks are not reinterpreted. Keep credentials in the host credential store and never echo them.
Do not browse for generic release advice when repository policy and previous comparable releases answer the question. Use official host documentation only for current API or CLI behavior.
test cases passed, vs 3 / 19 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.