Skill Installer Plus

Skill Installer Plus

Use this skill to streamline safe skill installs and continuously improve install decisions.

Workflow

1. Generate a recommendation plan from local candidates and prior outcomes.
2. Evaluate usage guardrails with $usage-watcher, $skill-cost-credit-governor, and $skill-cold-start-warm-path-optimizer before final admit decisions.
3. Admit selected skills through skill-arbiter in personal-lockdown mode.
4. Persist outcomes to the installer ledger and optionally ingest trust-ledger events.
5. Record manual post-install feedback after real usage.

Plan Recommendations

python3 "$CODEX_HOME/skills/skill-installer-plus/scripts/skill_installer_plus.py" \
  --ledger "$CODEX_HOME/skills/.skill-installer-plus-ledger.json" \
  plan \
  --skills-root skill-candidates \
  --dest "$CODEX_HOME/skills" \
  --trust-report /tmp/skill-trust-report.json \
  --json-out /tmp/skill-installer-plus-plan.json

Admit Skills

python3 "$CODEX_HOME/skills/skill-installer-plus/scripts/skill_installer_plus.py" \
  --ledger "$CODEX_HOME/skills/.skill-installer-plus-ledger.json" \
  admit \
  --skill skill-installer-plus \
  --source-dir skill-candidates \
  --dest "$CODEX_HOME/skills" \
  --window 10 --baseline-window 3 --threshold 3 --max-rg 3 \
  --arbiter-json /tmp/skill-installer-plus-arbiter.json \
  --json-out /tmp/skill-installer-plus-admit.json

Batch admit from a curated list:

python3 "$CODEX_HOME/skills/skill-installer-plus/scripts/skill_installer_plus.py" \
  --ledger "$CODEX_HOME/skills/.skill-installer-plus-ledger.json" \
  admit \
  --skills-file /tmp/candidate-skills.txt \
  --source-dir skill-candidates \
  --dest "$CODEX_HOME/skills" \
  --window 10 --baseline-window 3 --threshold 3 --max-rg 3 \
  --arbiter-json /tmp/skill-installer-plus-arbiter-batch.json \
  --json-out /tmp/skill-installer-plus-admit-batch.json

Record Manual Feedback

python3 "$CODEX_HOME/skills/skill-installer-plus/scripts/skill_installer_plus.py" \
  --ledger "$CODEX_HOME/skills/.skill-installer-plus-ledger.json" \
  feedback \
  --skill skill-installer-plus \
  --event success \
  --note "real workflow stable" \
  --json-out /tmp/skill-installer-plus-feedback.json

Show Ledger State

python3 "$CODEX_HOME/skills/skill-installer-plus/scripts/skill_installer_plus.py" \
  --ledger "$CODEX_HOME/skills/.skill-installer-plus-ledger.json" \
  show \
  --recent 10 \
  --json-out /tmp/skill-installer-plus-show.json

Scope Boundary

Use this skill only for the skill-installer-plus lane and workflow defined in this file and its references.

Do not use this skill for unrelated lanes; route those through $skill-hub and the most specific matching skill.

Admission decision is incomplete without usage guardrail evidence paths from usage-watcher, skill-cost-credit-governor, and skill-cold-start-warm-path-optimizer.

Evidence Contract

For each admit cycle, keep:

• plan_json
• admit_json
• arbiter_json
• trust_ingest_status
• optional feedback_json

If any required artifact is missing, installation governance is incomplete.

References

• references/learning-loop.md

Loopback

If this lane is unresolved, blocked, or ambiguous:

1. Capture current evidence and failure context.
2. Route back through $skill-hub for chain recalculation.
3. Resume only after the updated chain returns a deterministic next step.