operate — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited operate (Agent Skill) and scored it 96/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
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.
Run .skills/operate/scripts/operator_watch.sh from the project root. It watches a proxy you already started with the monitor skill (run_monitor.sh --detach) and emits one line per actionable event on stdout, staying quiet during healthy idle.
.skills/operate/scripts/operator_watch.sh # defaults below
.skills/operate/scripts/operator_watch.sh data/clawback.run.log data/clawback.run.pidDefaults match run_monitor.sh:
LOG_FILE = data/clawback.run.logPID_FILE = data/clawback.run.pidOPERATOR_WATCH_POLL_SEC (default 5s)[error] / [warn] / → 4xx|5xx / failed /self-loop / EADDRINUSE / ECONN* / unhandled / Error:
clawback listening (a (re)bind) andruntime mode toggled (e.g. the baseline→armed flip after a fresh proxy's 5-turn capture window, or any admin toggle)
[error] PROXY DOWN … line the moment the proxy pid stopsresponding — then it exits non-zero so the watch ends deliberately instead of hanging silent
run_monitor.sh --attach?--attach only tails the log. macOS/BSD tail has no --pid, so a crash makes the log go quiet and an attached follower can't tell "dead" from "idle". Only run_monitor.sh's foreground follow mode has a liveness watcher — but that mode starts its own proxy and would collide on the port with one already detached. This script adds liveness to the already-detached case without touching the port.
.skills/scripts/run_monitor.sh --detach — start the proxy..skills/operate/scripts/operator_watch.sh withpersistent: true. Each emitted line becomes a notification.
PROXY DOWN (or a crash signature), read the tail ofdata/clawback.run.log to find the cause — that's the "pattern in the logs that needs fixing" — then bring it back with the restart skill and re-arm the Monitor.
Stop the proxy with the shutdown skill; the watch exits on its own when the proxy is gone.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.