screenshot-redactor¶
Redact sensitive regions in a screenshot — full blur / solid box for high-risk (tokens, keys, cards, SSNs) and partial mask for low-risk illustrative data (emails, names) — using sharp or jimp, then VERIFY that every flagged region is actually obscured. If coverage can't be verified it BLOCKS the write. Produces only the redacted image; the raw stays gitignored.
At a glance¶
| Model | inherits the session's |
| Tools | Read Grep Glob Bash Write |
| Author | navjyotnishant |
| Source | agents/screenshot-redactor.md |
Writes files
This agent produces a file itself, rather than returning content for the skill to write — one of only two that do.
When it runs¶
The sensitive-data-reviewer returned marked regions on a dashboard screenshot.
redact the flagged regions
The capture-screenshots skill spawns this agent with the raw image + regions; it blurs, verifies coverage, and blocks the write if unsure.
What it returns¶
Return: the redacted image path (only on PASS), a per-region report (type · risk · style applied · verified yes/no), and the gate verdict. On BLOCK, no writable image — just the uncertain regions and recommended next step.
Spawned by¶
Derived, not declared
No agent file records which skills call it — this list is recovered from the skill definitions at build time, so it cannot go stale.
See the /capture-screenshots pipeline for where this fits in the whole run.