bridge-contribute — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited bridge-contribute (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
This is the skill the open-bridge community uses to feed features and improvements back: it turns work on your user/* branch into a clean, reviewed pull request against bks-lab/open-bridge (or your org's overlay repo) — and it refuses to ship anything the content-safety gate flags. Protecting your own customer names, personal data, and internal paths is a hard gate, not a checkbox.
Read the referenced file ONLY when triggered.
| Argument | Effect | Default |
|---|---|---|
(none) | Scan and categorize all contributable files | — |
{path} | Analyze a specific file or directory | — |
--adapt | Generalize org-specific content before contributing | false |
--repo <name> | Only consider candidates for one destination | all |
User wants to...
├── Scan branch for contributions → references/workflow.md
├── Contribute a specific file/dir → references/workflow.md ({path})
├── Adapt/generalize content first → references/workflow.md § Adapt mode
├── Promote whole commits by scope → use the bridge-promote skill (/promote)
└── Questions about scope routing → CLAUDE.md § tier modelNo PR is created until the two-layer gate passes for the exact outgoing file set:
python3 scripts/no-scrub-leak.py {files} — universal classes(absolute user paths, key/token shapes, merge-conflict markers) plus your own roster from scripts/leak-patterns-internal.txt (local-only, never shipped — you maintain your customer/PII regexes there; the shipped scanner cannot know them).
rules/promote-safety.md —the OSS upstream uses the strictest list.
REFUSE path: personal-PII or roster hits exclude the affected files — no override flag, no "it's just an example". Remediation: --adapt, re-scan, only a clean re-scan unblocks. Clean files in the same batch still ship.
/promote (bridge-promote skill) moves whole commits by scope: routing — the lightweight path when your branch has clean per-commit scope discipline. /contribute works file-level, with adaptation and always ending in a PR — the path for mixed branches and for first-time community contributions. Full comparison table: references/workflow.md § Relationship.
bridge-promote — commit-level scope-routed promote to upstreamsrules/contribute-advisor.md — proactive "this looks upstream-worthy" nudgesrules/promote-safety.md — the content-scan rules this skill enforcesCONTRIBUTING.md — repo-side contribution policy (DCO, licenses)~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.