What the reviewer found
Setup/ops skill for the Aeon scheduled-agent framework. Mode 8 reads the operator's own local Claude Code transcripts only to build an aggregate digest shown back to that same operator (explicitly documented as local-only, never inside an Aeon run, with a privacy note against pasting raw content elsewhere); the .mcp.json mention is troubleshooting prose, not a read. 'rm -rf aeon.git' removes a bare clone dir the same script just created. Installs a small CLI runtime and can grant write/PR access to the user's repo, so the person should know it manages GitHub secrets and can run with elevated s
installs-software— installs softwarenetwork— calls the vendor’s APIsettings-edit— edits agent settings
What was read
The file as it ships in aaronjmars/aeon-agent:
.claude/skills/aeon/SKILL.md
What the static scan said
The scan flagged 4things. The reviewer kept 0 and dismissed 4 as false.
AS1Reads agent configuration directories — false positiveAS2Reads MCP configuration — false positiveRMRecursive force delete — false positiveNETMakes network calls — false positive
How this review was made
Sonnet 5 read the files above on 7 September 2026 and answered three questions: is it dangerous to whoever installs it, is each scanner finding real, and what should the installer know. The verdict is bound to the file's hash; when the file changes, it is scanned afresh and reviewed again. A script that changes while the definition does not is not re-reviewed — that is a known gap. How the scan and the review work.