Escalate
AdvisoryAudited by Static analysis on Apr 30, 2026.
Overview
No suspicious patterns detected.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Saved rules may cause the agent to ask less often in matching future situations, so incorrect or stale entries could affect future decisions.
The skill stores persistent escalation rules, recent decisions, and domain overrides that future agents may reuse when deciding whether to ask or act.
Local state lives in `~/escalate/`: ... `~/escalate/memory.md` ... `~/escalate/decisions.md` ... `domains/`
Review ~/escalate/memory.md, decisions.md, and domain overrides periodically; keep secrets out and correct stale boundaries promptly.
Approved snippets could change how future agents in that workspace retrieve escalation memory and decide when to ask.
The skill can propose persistent workspace steering changes, but it requires showing the snippet and waiting for explicit approval before modifying workspace files.
If a workspace SOUL file exists... Show the exact snippet immediately and wait for explicit approval before writing it.
Approve AGENTS or SOUL edits only after reading the exact snippet and confirming it matches the workspace’s desired behavior.
First use may create local files in the user’s home directory for persistent escalation memory.
Setup includes simple local shell commands to create and permission the skill’s state directory; there is no evidence of downloads, network calls, or privileged paths.
mkdir -p ~/escalate/domains ... touch ~/escalate/{memory.md,decisions.md} ... chmod 700 ~/escalateAllow the setup commands only if you want this persistent local state; inspect the files afterward if desired.