exploring-scouts — Claude Code skill: what it does, how to install
exploring-scouts reads the state of PostHog Signals scouts: which ones a project has, how recent runs went, and where their reports ended up. Read-only.
no license 197 installs ★ 101 updated 2026-11-10
npx skills add PostHog/ai-plugin --skill exploring-scouts -a claude-codeWhat exploring-scouts does
This skill inspects the scouts of PostHog Signals, the agents that wake on a schedule, look at one project and file what they find into the Signals inbox. Handy when you want to know what a scout has been up to, or why a run came back with nothing.
Every call goes through read-only MCP tools. scout-config-list gives the roster and the schedules; from there you can open recent runs, walk through the reasoning of a single run, check the scratchpad memory the fleet keeps, and follow a run to the reports it wrote or edited. Health questions come off the same rows: how often it fires, how many runs succeed, how much noise it makes.
Nothing here writes or tunes a scout, that is what authoring-scouts covers. A zero in emitted_count is not an idle run: the column belongs to an older output channel and stays at zero for scouts that publish reports. If a scout-* name comes back unknown, look for the legacy signals-scout-* name in the tool catalog.
What it helps with
- see which scouts a project has and when each one fires
- find out why a run reported nothing
- follow a run to the reports it created or changed
- judge a scout's cadence and success rate over time
- read the memory the fleet built up between runs
Example prompts for the agent
- What are my scouts doing right now, and on what schedule?
- Walk me through the last run of my error-tracking scout and what it wrote
- Why did this scout find nothing on its last run?
How to install exploring-scouts
1Skills CLI (Vercel)
npx skills add PostHog/ai-plugin --skill exploring-scouts -a claude-code2Manually: put the files in place
~/.claude/skills/exploring-scouts/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/exploring-scouts/
1Skills CLI (Vercel)
npx skills add PostHog/ai-plugin --skill exploring-scouts -a codex2Manually: put the files in place
.agents/skills/exploring-scouts/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add PostHog/ai-plugin --skill exploring-scouts -a cursor2Manually: put the files in place
.cursor/skills/exploring-scouts/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install PostHog/ai-plugin skills/exploring-scouts2Manually: put the files in place
.github/skills/exploring-scouts/SKILL.md1Skills CLI (Vercel)
npx skills add PostHog/ai-plugin --skill exploring-scouts -a gemini-cli2Platform CLI
gemini skills install https://github.com/PostHog/ai-plugin --path skills/exploring-scoutsManually: put the files in place
.devin/skills/exploring-scouts/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add PostHog/ai-plugin --skill exploring-scouts -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about exploring-scouts
How do I install exploring-scouts in Claude Code?
Run npx skills add PostHog/ai-plugin --skill exploring-scouts -a claude-code. The skill file lands at ~/.claude/skills/exploring-scouts/SKILL.md.
What does exploring-scouts do?
It surveys a project's PostHog Signals scouts: schedules, recent runs, scratchpad memory and the reports they filed. All of it through read-only MCP tools, with no writes.
How is it different from authoring-scouts?
This one only reads and explains: what exists, how it ran, where the reports went. Writing a new scout or tuning an existing one belongs to authoring-scouts.