gtapps/hermit-evolution
plugins/claude-code-hermit/skills/hermit-evolution/SKILL.md
Show how this hermit has evolved: cost trend and source split, routines and watches, top things produced, and what grew since hatch. Activates on messages like 'how am I trending', 'cost trend', 'evolution report', 'am I improving', 'proposal velocity', 'monthly report', 'what did I produce last month'.
$ install --global
skillsauthnpx skillsauth add gtapps/claude-code-hermit hermit-evolutionInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Jul 28, 2026, 4:39 AM306.5s2 files scanned
SKILL.md
- name:
- hermit-evolution
- description:
- Show how this hermit has evolved: cost trend and source split, routines and watches, top things produced, and what grew since hatch. Activates on messages like 'how am I trending', 'cost trend', 'evolution report', 'am I improving', 'proposal velocity', 'monthly report', 'what did I produce last month'.
Hermit Evolution
Synthesize a coherent evolution report: cost trend and source split, autonomy and proposal velocity, active routines and watches, top things produced last month, and what grew organically since hatch.
Step 0 — Channel reply
If this skill was invoked from a channel-arrived message (the inbound prompt contains a <channel source="..."> tag), reply via that channel's reply tool. Otherwise emit to conversation.
Step 1 — Dispatch the eval runner
Dispatch claude-code-hermit:skill-eval-runner pointed at ${CLAUDE_PLUGIN_ROOT}/skills/hermit-evolution/reference.md. The runner reads the weekly review files, proposal metrics, config, session reports, and OPERATOR.md — and runs proposal.ts metrics and cost-report.ts reflect — in an isolated context, then returns the assembled evolution report. This keeps those heavy reads off this session's inherited context.
Pass plugin_root: ${CLAUDE_PLUGIN_ROOT} in the dispatch prompt — the runner reads reference.md as file content, where ${CLAUDE_PLUGIN_ROOT} is never substituted, so it needs the resolved absolute path to run the scripts and read template/skill paths.
Eval runner return schema — the runner's return value is a JSON object conforming to this block. The schema is byte-identical in reference.md (producer) and here (consumer); a contract test asserts this.
{
"report": "<assembled ≤1500-char report, section structure above>"
}
Failure policy: if the runner returns null or malformed JSON, fail-open — deliver a one-line "hermit-evolution: snapshot unavailable (analysis-runner failed)" via the Step 0 target and stop.
Step 2 — Deliver
Deliver the runner's report verbatim (≤1500 chars) via the Step 0 target. The runner already omits empty sections; do not re-synthesize. For reference, the report uses this section structure:
### Cost
- Trend (weekly): $A → $B → $C → $X (Δ +/-N%)
- Last 30d: $X.XX total — scheduled (routines/heartbeat): $Y (N%), interactive: $Z (N%)
### Autonomy
- Self-directed: N% this week (vs M% prior, Δ +/-N pp)
### Proposal velocity
- Median resolution: Nd (N proposals, last 30d)
[acceptance-by-source table, rows n>0 only; note triggered kill gates]
### Routines & watches
- heartbeat: every X
- <routine-id>: <cron> · fired N× in last 30d
- watch <monitor-id>: active
### Top-3 produced (inferred)
- <one-line description>
(inferred — no operator-used signal exists)
### Grown since hatch (approximated)
- OPERATOR.md: <sections added, or "no additions vs template">
- Skills: <project-local skill names, or "none">
(approximated — no hatch baseline stored)
Related Skills
gtapps/fitness-brief
tools
VerifiedTrustedCommunity
Composes and delivers the daily fitness brief — a forward-looking morning read (readiness + today's plan) or a backward-looking evening read (today's training, or an earned-rest note, + tomorrow's setup) — in the operator's configured voice. Invoke with /claude-code-fitness-hermit:fitness-brief --morning|--evening|--slot <name>. Becomes the plugin's two daily beats — the morning Strava connectivity check and the evening activity sync, RPE binding, and Run deep-dive.
68SKILL.mdUpdated Jul 25, 2026
gtapps/relogin
development
VerifiedTrustedCommunity
Renew the hermit's long-lived Claude login token over the channel, before it expires. Relays a one-time sign-in link to the operator, takes the code back, installs the new token, and restarts. Activates on messages like 'relogin', 'renew my login', 'reauth', 'the login is expiring', or when doctor's credential-expiry check flags setup-token.
68SKILL.mdUpdated Jul 22, 2026
gtapps/weekly-digest
development
VerifiedTrustedCommunity
Synthesizes the past 7 days of archived briefs into a weekly digest — top stories, emerging vs faded themes, category activity, and per-source performance built from archive frontmatter. Delivers to the operator's configured channel and archives a weekly note. Designed as a weekly routine. Invoke with /feed-hermit:weekly-digest.
68SKILL.mdUpdated Jul 21, 2026
gtapps/story-arcs
development
VerifiedTrustedCommunity
Manage developing story arcs tracked across briefs — add, resolve, and list active arcs in compiled/story-arcs-*.md. Arc Watch keywords drive the feed-brief arc-tagging enrichment. Invoke with /feed-hermit:story-arcs add|resolve|list.
68SKILL.mdUpdated Jul 21, 2026