Enforce Redaction Gate

A redaction gate is the privacy-boundary analogue of an AST checker: it parses what is about to ship and refuses to let a deny-listed shape through, returning a machine-checkable verdict rather than a human opinion. This skill builds that gate as a single executable that any transform skill or CI job can call to verify its output. Two tiers cooperate — a fast shape-based deny-list and a slower structure-aware pass — so the gate catches both literal token leaks and the structural leaks a line-grep misses.

When to Use

• Gating a public mirror or disclosure draft before publish (the blocking check, not a warning)
• Verifying the output of a redaction transform (redact-visualization-for-disclosure, redact-wire-capture) actually cleared every pattern
• Adding a CI step that fails the build when a sensitive identifier shape appears
• Replacing an ad-hoc grep -r secret with a maintained, exit-code-contracted, structure-aware checker
• Auditing an already-published tree after a near-miss to confirm nothing slipped through

Inputs

• Required: A target tree or file set to scan (the artifact about to ship)
• Required: A deny-list of sensitive identifier shapes (regexes), kept in the private source repo — never published
• Optional: A structure map — which file types carry structured content (JSON keys, HTML attributes, SVG text nodes, code identifiers) and which positions are sensitive
• Optional: A CI context (the gate runs the same locally and in CI)