Carol — back to Apps ← Apps

Carolopedia

A friendly guide to Carol, her ecosystem, and the agents who built her.

📖 CarolopediaServicesBuild InitiativesAll activitiesINI-999900321
📋

CAROL-INI-2088-00: Upgrade existing skills to glean Carolverse patterns + nightly standards sweep

Initiative
Open in Initiatives →

📖About

Follow-on to CAROL-INI-2087. NEW skills now glean existing implementations, but the ~35 EXISTING skills do not (only 3 app-skills reference any standard). Give every skill a pattern-standards checklist of the existing Carolverse patterns for its activity, auto-derived by the miner, and keep them current via a nightly sweep. SCOPE: (1) shared/skill_patterns.py: per-activity keyword/statement resolution (curated overrides + auto-derive from skill name+description), generate_for_skill (runs skill_pattern_miner, writes /patterns.md with a machine-readable fingerprint for drift detection), checklist_for_skill / checklist_for_active_step (for injection), sweep_all. (2) Generate patterns.md for all ~35 activity skills (initial population). (3) AUTO-INJECT the active skill pattern checklist into the build context for ALL skills: extend Sage spec stage + Archon design stage to bake carolverse_skill_patterns into the design_spec + dev_spec (mirrors the design_standards injection from 2019, not gated to app work) so Forge always sees the proven pattern. (4) RECIPE REFERENCE: add a uniform consult-your-patterns.md line to each skill -execute and -design recipe. (5) NIGHTLY SWEEP: a Sage-owned scheduled droid (cron + create_run/update_run run-audit per building-scheduled-processes), regenerates all patterns.md, auto-updates on change, and files ONE consolidated planner review initiative per run when at least one skill pattern MATERIALLY drifted (fingerprint change), flags a skill that had patterns but now mines nothing as possible breakage. (6) Make create-a-new-skill also write a patterns.md for new skills (close the loop). (7) cookbook note + regression. Owner of skills stays Sage.

⚖️Decisions

  • Auto-detected remediation target INI-999900320 from title/description scan (matched CAROL-INI-2087 -> row id 999900320 (CAROL-INI-2087-00: Skill authoring mines existing implementations to glean the p)); override by setting remediates_initiative_id explicitly at bypass_start. (system-auto-detect)
  • Elrond's bypass methodology checklist (a reminder, not a gate -- you've got this): 0. File it requested_mode='bypass' (planner-vs-bypass is a deliberate choice). bypass_start REFUSES a non-bypass initiative (CAROL-INI-1846), and the dispatcher only skips the bypass lane when the mode says bypass -- a 'planner' mistag lets Merlin's pipeline grab the placeholder step and block your finished work. 1. Filed as planned status -- let the bypass claim/activate it; never file active. 2. Open the bypass (bypass_start) with your droid id + the remediation answer (remediates_initiative_id=NNN, or remediates_nothing=True). 3. Work the blocks for your work-type: template -> design -> code -> test -> review. Do the real work; record decisions on the initiative as you make them. 4. Reality is recorded for you at close -- code (files changed), each decision, and the twin-review verdict become real activities tied to this initiative and show in the Activity Tracker like a planner run (CAROL-INI-1840). No dummy rows. 5. Keep the initiative status moving; it parks in 'reviewing' and is tagged uat-pending for you at close (CAROL-INI-1836), so the stuck-watchdog leaves it alone until UAT. 6. Close runs the gates (design/architecture compliance + caller-audit). If a gate flags something pre-existing or unrelated to your change, waive it with a clear written rationale -- audit, don't skip. 7. Bypass skips the planner's auto-orchestration, NOT the standards. Same template checklist, same review, same observability as a planner run. (elrond)
  • [status-router] planned -> executing | event=bypass_executing | bypass transition (or-bx-01)
  • [status-router] executing -> reviewing | event=bypass_reviewing | bypass transition (or-bx-01)
  • [status-router] reviewing -> closed | event=operator_signoff | Auto-accepted (CAROL-INI-1859): Orion-initiated, >2 days in reviewing with no objection. (el-srac-01)

Success criteria

  • Every activity skill has a patterns.md checklist of existing Carolverse patterns for its activity, auto-derived by the miner, with a machine-readable fingerprint (must_have)
  • shared/skill_patterns.py exists with generate_for_skill, checklist_for_skill, checklist_for_active_step, and sweep_all; importable and callable (must_have)
  • The active skill pattern checklist is auto-injected into the build context for all skills (Sage spec + Archon design stages bake carolverse_skill_patterns into design and dev specs) (must_have)
  • Each skill -execute and -design recipe references its patterns.md (must_have)
  • A Sage-owned scheduled sweep droid regenerates patterns nightly with run-audit, auto-updates on change, and files one consolidated review initiative on material drift (must_have)
  • create-a-new-skill writes a patterns.md for new skills; cookbook note + regression added (must_have)