Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Orchestrate end-to-end release pipelines. Use when running pre-release validation, generating changelogs, bumping semantic versions, scoring deployment readiness, or gating releases with secret scanning and GO/NO-GO checks.
.claude/skills/borghei-release-orchestrator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 560% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 68% | 0% |
The agent runs pre-flight validation, generates changelogs from conventional commits, auto-bumps semantic versions, and scores deployment readiness with a GO/CONDITIONAL/NO-GO decision.
package.json, pyproject.toml, Cargo.toml, etc.; pre-release tags (--pre alpha|beta|rc).@author attribution.Before orchestrating the release, confirm these inputs. If any is unknown or vague, ASK — do not assume:
--from/--to)Stop rule: ask only the 2-3 that most change the output. If the user says "just draft it," proceed and list your assumptions at the top of the artifact.
| Tool | Purpose | Command | |------|---------|---------| | preflight_checker.py | Run 7 pre-release checks (sync, conflicts, secrets, commits, deps) | python scripts/preflight_checker.py --repo . --base main --json | | changelog_generator.py | Generate Keep a Changelog markdown from a ref range | python scripts/changelog_generator.py --repo . --from v1.2.0 --to HEAD --output CHANGELOG.md | | version_bumper.py | Auto-detect next semver from commits; write version files | python scripts/version_bumper.py --repo . --dry-run --json | | release_readiness_scorer.py | Score readiness 0-100 with GO/CONDITIONAL/NO-GO decision | python scripts/release_readiness_scorer.py --input release_data.json --json |
All tools support --json for machine output. Exit code 0 = pass, 1 = fail (CI-friendly).
Load the reference that matches the task — keep this file lean and pull detail on demand:
This skill covers: pre-flight validation, semantic version bumping, changelog generation, deployment-readiness scoring, and gating releases with GO/NO-GO decisions for git + conventional-commits projects.
This skill does NOT cover: the actual deploy/orchestration execution (handled by senior-devops / devops-workflow-engineer), test authoring (senior-qa), or deep security scanning beyond secret pattern-matching (senior-secops).
| Skill | Integration | |-------|-------------| | senior-devops | Pipeline stages consume pre-flight and readiness JSON as gates | | senior-qa | Test results feed Tests category (25% weight) | | senior-secops | Secret scan and CVE counts feed Security category (15%) | | code-reviewer | Code quality metrics feed Code Quality category (20%) | | devops-workflow-engineer | Workflow YAML calls tools as pipeline steps |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-06 | fail→pass | 10,888 | 5,508 | -49% | 1 | 1 | 0% | 1,835 | 2,212 | +21% | 0 | 0 | — |
case-01 | fail→fail | 8,316 | 20,715 | +149% | 1 | 1 | 0% | 344 | 4,956 | +1341% | 0 | 0 | — |
case-02 | fail→pass | 4,155 | 11,208 | +170% | 1 | 1 | 0% | 499 | 3,293 | +560% | 0 | 0 | — |
case-03 | fail→pass | 14,830 | 13,479 | -9% | 1 | 1 | 0% | 2,313 | 3,685 | +59% | 0 | 0 | — |
case-04 | fail→pass | 10,872 | 2,453 | -77% | 1 | 1 | 0% | 1,531 | 1,699 | +11% | 0 | 0 | — |
case-05 | fail→pass | 6,605 | 2,821 | -57% | 1 | 1 | 0% | 1,054 | 1,775 | +68% | 0 | 0 | — |
case-07 | fail→pass | 19,416 | 2,774 | -86% | 1 | 1 | 0% | 762 | 1,723 | +126% | 0 | 0 | — |
case-08 | pass→pass | 6,566 | 2,322 | -65% | 1 | 1 | 0% | 1,006 | 1,632 | +62% | 0 | 0 | — |
case-09 | pass→pass | 10,380 | 4,074 | -61% | 1 | 1 | 0% | 1,576 | 1,943 | +23% | 0 | 0 | — |
case-10 | pass→pass | 7,977 | 3,733 | -53% | 1 | 1 | 0% | 1,133 | 1,918 | +69% | 0 | 0 | — |
case-11 | pass→pass | 19,944 | 7,632 | -62% | 1 | 1 | 0% | 3,801 | 2,502 | -34% | 0 | 0 | — |
case-12 | pass→pass | 10,494 | 2,176 | -79% | 1 | 1 | 0% | 1,514 | 1,599 | +6% | 0 | 0 | — |
case-13 | fail→pass | 8,036 | 2,938 | -63% | 1 | 1 | 0% | 1,248 | 1,793 | +44% | 0 | 0 | — |
case-14 | fail→pass | 7,125 | 2,661 | -63% | 1 | 1 | 0% | 1,221 | 1,741 | +43% | 0 | 0 | — |
case-15 | pass→pass | 12,824 | 2,151 | -83% | 1 | 1 | 0% | 1,842 | 1,610 | -13% | 0 | 0 | — |
case-16 | pass→pass | 10,830 | 2,794 | -74% | 1 | 1 | 0% | 1,583 | 1,716 | +8% | 0 | 0 | — |
case-17 | fail→pass | 10,014 | 1,888 | -81% | 1 | 1 | 0% | 1,525 | 1,519 | -0% | 0 | 0 | — |
case-18 | fail→pass | 8,405 | 3,126 | -63% | 1 | 1 | 0% | 1,317 | 1,818 | +38% | 0 | 0 | — |
case-19 | fail→pass | 11,687 | 2,408 | -79% | 1 | 1 | 0% | 1,754 | 1,639 | -7% | 0 | 0 | — |
case-20 | pass→pass | 4,657 | 7,166 | +54% | 1 | 1 | 0% | 716 | 2,446 | +242% | 0 | 0 | — |
case-21 | fail→fail | 18,909 | 14,007 | -26% | 1 | 1 | 0% | 3,620 | 4,038 | +12% | 0 | 0 | — |
case-22 | fail→pass | 16,654 | 15,580 | -6% | 1 | 1 | 0% | 2,225 | 2,817 | +27% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +55 percentage points is the difference between those two pass rates over the 21 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.