Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/OSideMedia/higgsfield-ai-prompt-skillWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/osidemedia/higgsfield-ai-prompt-skill/release)<a href="https://agentmods.dev/commands/osidemedia/higgsfield-ai-prompt-skill/release"><img src="https://agentmods.dev/badge/commands/osidemedia/higgsfield-ai-prompt-skill/release/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/osidemedia/higgsfield-ai-prompt-skill/release"><img src="https://agentmods.dev/badge/commands/osidemedia/higgsfield-ai-prompt-skill/release.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00024 | $0.01018 |
| Opus 5.5 | $0.00010 | $0.00407 |
| Sonnet 5.5 | $0.00005 | $0.00204 |
| Haiku 4.5 | $0.00002 | $0.00102 |
Grade A, and why
release scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Walk through a release for version $ARGUMENTS (e.g. /release 2.1.0).
If no version argument is provided, check CHANGELOG.md for the latest version and ask what the new version should be.
main is protected: nothing is committed or pushed to it directly. The release
lands through a PR, and the tag goes on the merge commit that PR creates on
main — never on the branch commit. (CLAUDE.md § Rules is the source of this
ceremony; keep the two in step.)
Steps:
- Sanitize the version argument — before using
$ARGUMENTSin any shell command, confirm it is a bare semantic version matching^[0-9]+\.[0-9]+\.[0-9]+$(e.g.2.1.0). If it contains anything else (spaces,;, backticks, quotes, path separators), STOP and ask the user for a clean version string — never interpolate an unvalidated argument into thegit/ghshell snippets below. - Release branch — from an up-to-date
main:git switch main && git pull --ff-only && git switch -c release/v$ARGUMENTS(or continue on the existingrelease/v$ARGUMENTS). Never commit onmain. - Bump + changelog — set the root
SKILL.mdfrontmattermetadata.version/metadata.updatedand the README badge + footer (validate.py checks they agree); confirm CHANGELOG.md has av$ARGUMENTSentry — if not, ask what to add. - Release gate —
python3 scripts/validate.py --strict,python3 -m pytest tests/ -q,python3 scripts/validate.py --evals. Stop if any of the three fails. - User guide: regenerate → refresh MANIFEST — AFTER the version bump:
python3 scripts/generate_user_guide.py, thenpython3 scripts/validate_user_guide.py(manifest comparison; review any flagged drift), thenpython3 scripts/validate_user_guide.py --write-manifestand stagedocs/user-guide/MANIFEST.json. The PDF is a release artifact — git-ignored, never committed; keep the generateddocs/user-guide/USER-GUIDE.pdffor step 9. - Commit on the branch —
feat: v$ARGUMENTS — <summary from changelog>(orfix:for a fix release). Re-run the three gates if anything changed after step 3. - Push the branch + open the PR — confirm with the user first:
git push -u origin release/v$ARGUMENTS, thengh pr create --base main --head release/v$ARGUMENTS --title "v$ARGUMENTS — <summary>" --body-file -with the changelog entry. Wait for thevalidateCI check to pass. - Merge — confirm with the user; merge the PR with a merge commit (
gh pr merge <number> --merge). Then read the merge commit id from GitHub, not from local history:MERGE=$(gh pr view <number> --json mergeCommit --jq .mergeCommit.oid). - Tag the MERGE commit —
git fetch origin && git tag -a v$ARGUMENTS "$MERGE" -m "v$ARGUMENTS"and confirmgit rev-parse v$ARGUMENTS^{commit}equals$MERGEand is onorigin/main(git merge-base --is-ancestor "$MERGE" origin/main). Push only the tag:git push origin v$ARGUMENTS. - GitHub release + PDF — confirm with the user:
gh release create v$ARGUMENTS --verify-tag --title "v$ARGUMENTS" --notes-file -using the changelog entry, thengh release upload v$ARGUMENTS docs/user-guide/USER-GUIDE.pdf. If anything else merged intomainbetween step 4 and the merge, check out the tag, regenerate the PDF and confirmpython3 scripts/validate_user_guide.pymatches the committed MANIFEST before uploading. - Clean up — delete the release branch on both sides after the merge:
git push origin --delete release/v$ARGUMENTSandgit branch -d release/v$ARGUMENTS.
Confirm with the user before each visible step (push, PR, merge, tag push, release, upload). Never force-push, never tag a branch commit, never commit to main.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago Changed · +7 lines · +10 tokens per session 453a1d88ed6e
- 1mo ago First seen · 22 lines · 14 tokens per session scan A b87df6c29a2a
release is a command published in the GitHub repository OSideMedia/higgsfield-ai-prompt-skill (809 stars, last pushed 3d ago), licensed MIT. It adds 24 tokens to every session and 1,018 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
land
Resolve and squash merge one named, currently authorized pull request.
quality
A command for temporarily clearing a release quality gate or one named check, while recording the reason and expiry time. A quality gate is a required check that normally blocks completion or release.
release-notes
Generate consistent, well-structured release notes from git history. Triggered on release tags following semver patterns (v..) to produce categorized changelog with breaking changes, features, fixes, and contributor attribution.
release-plan
A release-planning command for a software change. It documents the release scope, rollback plan, gradual or full rollout, and monitoring before waiting for human approval.
release-manager
Automated release coordination and deployment with ruv-swarm orchestration for seamless version management, testing, and deployment across multiple packages.
doc-changelog
Generation and maintenance of the project changelog.