config.example.json

{
  "schema_version": "1.5.0",
  "project": {
    "name": "my-app",
    "locale": "en",
    "timezone": "UTC"
  },
  "mission": "One sentence: what this project is for and what 'done well' means. The loop drives toward this. Set by /ooda-setup; empty string disables mission-aware scoring.",
  "domains": {
    "service_health": {
      "weight": 2.0,
      "state_file": "agent/state/service_health.json",
      "primary_skill": "/scan-health",
      "chain": [],
      "branch_prefix": "auto/scan-health/",
      "fallback": true,
      "enabled": true,
      "status": "active",
      "mission_alignment": 1.0
    },
    "test_coverage": {
      "weight": 0.5,
      "state_file": "agent/state/test_coverage.json",
      "primary_skill": "/check-tests",
      "chain": [],
      "branch_prefix": "auto/check-tests/",
      "fallback": true,
      "enabled": true,
      "status": "active",
      "mission_alignment": 1.0
    },
    "backlog": {
      "weight": 0.3,
      "state_file": "agent/state/backlog.json",
      "primary_skill": "/plan-backlog",
      "chain": [],
      "branch_prefix": "auto/plan-backlog/",
      "fallback": false,
      "enabled": true,
      "status": "active",
      "mission_alignment": 0.5
    },
    "business_strategy": {
      "weight": 1.0,
      "state_file": "agent/state/business_strategy.json",
      "primary_skill": "/scan-market",
      "chain": [],
      "branch_prefix": "auto/scan-market/",
      "fallback": true,
      "enabled": true,
      "status": "available",
      "description": "Market research and strategic analysis. Run /ooda-skill create scan-market to configure.",
      "mission_alignment": 0.5
    },
    "ux_evolution": {
      "weight": 1.0,
      "state_file": "agent/state/ux_evolution.json",
      "primary_skill": "/scan-ux",
      "chain": [],
      "branch_prefix": "auto/scan-ux/",
      "fallback": true,
      "enabled": true,
      "status": "available",
      "description": "UX audit and UI analysis. Run /ooda-skill create scan-ux to configure.",
      "mission_alignment": 0.5
    },
    "competitors": {
      "weight": 0.3,
      "state_file": "agent/state/competitors.json",
      "primary_skill": "/scan-competitors",
      "chain": [],
      "branch_prefix": "auto/scan-competitors/",
      "fallback": false,
      "enabled": true,
      "status": "available",
      "description": "Competitor monitoring and intelligence. Run /ooda-skill create scan-competitors to configure.",
      "mission_alignment": 0.5
    }
  },
  "domain_dependencies": {},
  "implementation": {
    "enabled": false,
    "weight": 1.5,
    "primary_skill": "/dev-cycle",
    "observe_loop_escape_bonus": 5.0,
    "mission_alignment": 1.0
  },
  "scoring": {
    "staleness_curve": "logarithmic",
    "balance_enabled": true,
    "balance_weight": 5.0,
    "goal_weight": 0.3,
    "confidence_weight": 0.2,
    "hours_if_never_run": 168,
    "implementation_formula": {
      "pending_multiplier": 2.0,
      "age_multiplier": 0.1,
      "rice_multiplier": 0.01,
      "open_pr_penalty": -3.0
    },
    "rice_extensions": {},
    "__rice_extensions_doc__": "Optional per-label RICE multipliers applied during action extraction (evolve Step 5), e.g. {\"security\": 2.0, \"chore\": 0.5}. Empty = no extensions.",
    "mission_weight": 6.0,
    "dry_domain_dampen": 0.3,
    "monitor_dry_dampen": 0.6,
    "off_mission_threshold": 0.2,
    "off_mission_dampen": 0.2
  },
  "confidence": {
    "initial": 0.7,
    "merge_boost": 0.1,
    "reject_penalty": 0.2,
    "min": 0.1,
    "max": 1.0,
    "observation_micro_adjustments": true
  },
  "safety": {
    "halt_file": "agent/safety/HALT",
    "confidence_threshold": 0.6,
    "min_cycle_interval_minutes": 30,
    "max_prs_per_cycle": 1,
    "first_cycle_observe_only": true,
    "max_files_per_pr": 20,
    "max_lines_per_pr": 500,
    "protected_paths": [
      "agent/safety/*",
      "skills/evolve/*",
      "agent/contracts/*"
    ],
    "skill_allowlist": [
      "/scan-health",
      "/check-tests",
      "/plan-backlog",
      "/run-deploy",
      "/dev-cycle",
      "/scan-market",
      "/scan-ux",
      "/scan-competitors"
    ],
    "lock_timeout_minutes": 30,
    "max_chain_depth": 3,
    "max_silent_failures": 3,
    "generated_file_patterns": [],
    "enable_rollback": false,
    "enable_auto_merge": false,
    "auto_merge_max_files": 5,
    "auto_merge_max_lines": 100,
    "__auto_merge_doc__": "enable_auto_merge (default false) is the SINGLE opt-in switch for autonomous merging. When true AND level>=3, evolve auto-merges ONLY low-risk PRs: no protected paths, files <= auto_merge_max_files, lines <= auto_merge_max_lines, and tests green. Everything else stays a Draft PR you merge by hand. Auto-merge takes a pre-action checkpoint and auto-reverts + HALTs on a failed post-merge health check (it forces enable_rollback behavior). Off by default: you stay in command unless you flip this.",
    "risk_rules": []
  },
  "progressive_complexity": {
    "current_level": 0,
    "levels": {
      "0": {
        "name": "Just watching",
        "domains": 1,
        "implementation": false,
        "auto_merge": false
      },
      "1": {
        "name": "Watching + testing",
        "domains": 2,
        "implementation": false,
        "auto_merge": false
      },
      "2": {
        "name": "Full observation",
        "domains": "all",
        "implementation": false,
        "auto_merge": false
      },
      "3": {
        "name": "Autonomous",
        "domains": "all",
        "implementation": true,
        "auto_merge": false,
        "__auto_merge_doc__": "Descriptive only \u2014 the engine reads config.safety.enable_auto_merge (NOT this field) to decide auto-merge. Level 3 enables autonomous implementation (Draft PRs by default); set safety.enable_auto_merge=true to also auto-merge low-risk PRs. See safety.__auto_merge_doc__ and README 'Auto-merge status'."
      }
    },
    "last_level_change": null
  },
  "signals": {
    "health_alert_bonus": 5.0,
    "alert_cooldown_hours": 4,
    "max_consecutive_alert_cycles": 3,
    "stale_after_change_hours": 24,
    "stale_after_change_bonus": 2.0,
    "queue_pressure_threshold": 5,
    "queue_pressure_bonus": 3.0,
    "queue_age_hours": 48,
    "queue_age_bonus": 2.0
  },
  "saturation": {
    "warn_threshold": 5,
    "boost_threshold": 10,
    "halt_threshold": 15,
    "auto_halt": true,
    "implementation_boost": 5.0
  },
  "memory": {
    "working_memory_size": 20,
    "episode_retention_weeks": 52,
    "contrarian_check_interval": 10,
    "action_queue_decay_days": 14,
    "action_queue_decay_amount": 0.05,
    "principle_similarity_threshold": 0.5,
    "principle_min_occurrences": 2,
    "reflection_recall_count": 3,
    "reflections_buffer_size": 20,
    "outcomes_buffer_size": 200
  },
  "season_modes": {
    "enabled": false,
    "current_mode": "default",
    "modes": {
      "default": {
        "weight_overrides": {},
        "disabled_domains": [],
        "signal_bonuses": {}
      }
    },
    "__doc__": "Toggle via /ooda-config season <mode>. When enabled: Step 1-A applies weight_overrides to domain_config.weight in-memory; disabled_domains are filtered out alongside status=disabled; signal_bonuses are merged into the cycle's signal table in Step 3-B. The 'default' mode acts as the static multiplier case (e.g., {\"weight_overrides\":{\"service_health\":2.0}} encodes a permanent 2x). See CONCEPTS.md."
  },
  "active_context": {
    "path": null,
    "refresh_skill": null,
    "refresh_interval_hours": 168,
    "__doc__": "Optional stakeholder context blob loaded by evolve Step 1 and exposed to every invoked skill as a context variable. The blob is opaque; each skill interprets it domain-specifically. If refresh_skill is set and the file is older than refresh_interval_hours, evolve queues a refresh cycle. Read-only by default."
  },
  "output": {
    "cycle_card": true,
    "__doc__": "When true (default), evolve renders a shareable Cycle Card at the end of every cycle (Step 7) summarizing Observe/Orient/Decide/Act, the single thing it Learned (re-aimed) this cycle, and the cost. Re-render the latest card any time with /ooda-status --share. Set false to suppress."
  },
  "test_command": "npm test",
  "test_timeout_seconds": 300,
  "health_endpoints": [],
  "health_check_timeout_seconds": 10,
  "deploy_workflow": null,
  "deploy_workflow_inputs": {},
  "deploy_monitor_timeout_seconds": 600,
  "deploy_health_wait_seconds": 30,
  "notifications": {
    "telegram": {
      "enabled": false,
      "bot_token": "$OODA_TELEGRAM_BOT_TOKEN",
      "chat_id": "$OODA_TELEGRAM_CHAT_ID"
    }
  },
  "cost": {
    "daily_limit_usd": 10.0,
    "warning_threshold_pct": 80,
    "max_backfill_cycles": 100,
    "__doc__": "daily_limit_usd: crossing it auto-creates a HALT (hard stop). warning_threshold_pct: warn at this % of the cap. max_backfill_cycles: Step 6-C8 caps how many missing ledger entries it will synthesize in one cycle so a corrupt cycle counter can't write thousands of entries (default 100)."
  },
  "eval": {
    "enabled": false,
    "model": "claude-haiku-4-5",
    "grade_on": [
      "pr_created",
      "pr_merged",
      "action_extracted"
    ],
    "__doc__": "Opt-in maker/checker layer (evolve Step 7-B). When enabled, a SEPARATE model independently judges whether each gradeable cycle achieved its declared GOAL (goal-conformance) and writes {achieved,reason,confidence} into outcomes.json verifier_verdict. Never overrides the score — a second opinion. Distinct from quality_rubric/Step 5-G, which judges whether the ARTIFACT is good and DOES move the score. Default off = zero extra cost."
  },
  "quality_rubric": {
    "__doc__": "ARTIFACT-quality axis (evolve Step 5-G, v1.7.0) — the fix for the dogfood failure where every cycle scored 0.5 / graded A while the built thing was broken, because nothing measured the artifact. CANONICAL placement is PER-DOMAIN: config.domains[<build domain>].quality_rubric (evolve is domain-agnostic). This top-level block is the single-domain fallback. Each cycle that produces an artifact, an INDEPENDENT critic (separate model context) captures the real artifact via capture_method and scores each dimension 0..1; rubric_score.py aggregates to artifact_score, which MULTIPLIES the process score in 6-C9 and drives the Goodhart Guard + LEAP trigger. The rubric is HUMAN-AUTHORED and read-only to the loop — add 'quality_rubric' / the config path to safety.protected_paths so the loop can never write its own grading standard (gaming-resistance). Empty dimensions = artifact axis OFF (process-only scoring, back-compat).",
    "bar": 0.65,
    "__bars_doc__": "v1.9.0 'Ambition' — DUAL thresholds break the prototype plateau. bar_leap: below it, ALWAYS leap. bar_coast: only above it may the loop coast (stop leaping). The forcing zone (bar_leap..bar_coast) keeps leaping on stagnation so the loop cannot declare victory at prototype quality. Set bar_coast HIGH (~0.85) anchored to a REAL product, not a demo. Legacy single `bar` → bar_leap==bar_coast==bar (old behaviour).",
    "bar_leap": 0.65,
    "bar_coast": 0.85,
    "prototype_ceiling": 0.20,
    "capture_method": "screenshot",
    "capture_command": "<serve + screenshot for web UIs | run + capture stdout for api_call | run benchmark>",
    "plateau_window": 4,
    "plateau_eps": 0.05,
    "locked": true,
    "__anchors_doc__": "v1.9.0 — each dimension SHOULD carry `reference` anchors naming what score_0.10/0.40/0.70/0.90 look like in REAL products, and `ceiling_without_assets`. The 5-G critic scores against these named anchors, not relative to the artifact's own past — without them a critic silently grades a flat-shaded prototype as 0.6 ('it works') instead of 0.1 ('vs Gran Turismo'). When artifact_score >= a dimension's ceiling_without_assets, the loop records a `human_required` skill_gap (authored model/track/audio assets needed) instead of leaping fruitlessly. Reference anchors + ceilings are human-authored + protected.",
    "__dimensions_doc__": "v1.8.0: each dimension may override capture_method so the critic gets the evidence it actually needs. 'screenshot' axes share one capture; EXPERIENTIAL axes (feel/fun/responsiveness) a screenshot cannot judge use 'gameplay_metrics' — a HUMAN-AUTHORED harness that exercises the artifact and emits metrics JSON. The harness MUST be in safety.protected_paths AND match gameplay_metrics_hash (independence gate, same invariant as the rubric hash); else the dimension scores null (capture_failure) rather than faking a score. Without per-dimension capture, experiential axes freeze at their initial score and silently cap artifact_quality. v1.8.1 rule (validated by the f1 probe): the harness must MEASURE BEHAVIOUR (e.g. drive the real physics and read the resulting numbers), NOT assert an implementation fact — a hardcoded flag like {feature: false} cannot credit a real fix, so it would trigger a spurious thrashing-HALT. Drive the artifact and report what it actually does.",
    "dimensions": [],
    "__example_dimension__": {
      "name": "visual_fidelity",
      "weight": 0.25,
      "capture_method": "screenshot",
      "__capture_states_doc__": "v1.12.0 (feedback fidelity) — a dimension is only gradeable in the state(s) where it MANIFESTS; a single default frame under-credits + misdirects the loop. List the states to drive the artifact into before capturing; the critic gets ALL frames. The f1 probe earned this: a chase camera / HDRI sky / speed blur / slip-angle physics were all added but every capture was one low-speed cockpit frame, so the work was invisible and the grade stalled. A state that can't be reached = capture_failure (null), not a low score.",
      "capture_states": ["chase view, car stationary on a straight", "chase view at ~80% top speed", "camera pitched up to show the sky/horizon"],
      "description": "3D visual quality vs SHIPPED games. Score against the reference anchors, not the artifact's past.",
      "reference": {
        "score_0.10": "flat-shaded primitive meshes, solid-colour sky, no shadows/post-processing (a 1990s look)",
        "score_0.40": "textured surfaces, basic shadows, simple post (PS2-era)",
        "score_0.70": "PBR materials + image-based lighting, bloom + tone mapping, particle FX, real skybox",
        "score_0.90": "authored models, motion blur, SSAO, weather, near-photoreal"
      },
      "ceiling_without_assets": 0.35,
      "ceiling_note": "Procedural geometry + CDN Three.js addons (post-processing, PBR/IBL, particles, Sky shader) top out ~0.35. Above that needs authored glTF models / textures / HDRIs — the loop records a human_required skill_gap, not keep leaping.",
      "__asset_sources_doc__": "v1.10.0 hand-off — the OPERATOR supplies authored assets here (CDN URLs or repo paths to glTF models, textures, HDRIs, audio) once the code-only ceiling is hit. Non-empty asset_sources LIFTS ceiling_without_assets (rubric_score.asset_ceiling → None): the human_required gap resolves and the loop resumes leaping toward bar_coast, now integrating the assets. The f1 probe drove this: the loop flagged the box-car ceiling at ~0.21; a real glTF car + an HDRI were supplied here and leaping resumed.",
      "asset_sources": [],
      "techniques": [
        "EffectComposer: RenderPass + UnrealBloomPass + OutputPass (ACES tone map)",
        "PMREMGenerator from RoomEnvironment or Sky → scene.environment (IBL)",
        "MeshStandardMaterial roughness/metalness + envMapIntensity (PBR paint)",
        "three/addons Sky shader (atmospheric scattering) replacing solid clear colour",
        "Points/BufferGeometry particle systems (tyre smoke, sparks)",
        "procedural normalMap (canvas noise) on the road/track material"
      ],
      "technique_cdns": ["three/addons/postprocessing/*", "three/addons/environments/RoomEnvironment.js", "three/addons/objects/Sky.js"],
      "__experiential_example__": {
        "name": "driving_feel", "weight": 0.25, "capture_method": "gameplay_metrics",
        "gameplay_metrics_command": "node tools/feel_harness.mjs",
        "gameplay_metrics_hash": "<sha256 of the harness file>",
        "metrics_fields": ["steer_response_rad_s", "has_oversteer", "has_weight_transfer"],
        "description": "Responsive, believable handling — judge metrics vs targets, not vibes."
      }
    }
  },
  "references": {
    "__doc__": "v1.11.0 (anti-maze) — REFERENCE TARGETS that ground the loop in external ground truth instead of model priors. dev-cycle Step 3-PRE resolves the reference for a technique, WebFetches the concrete block, derives acceptance criteria, THEN generates (AlphaCodium pattern). The 5-G critic scores AGAINST these (not the artifact's own past). Permanent: also mirror into agent/state/evolve/principles.json so they survive episode rollover. Populate per project with named real-product levels and reference-implementation URLs.",
    "visual": "https://threejs.org/examples/webgl_materials_car.html",
    "physics": "https://github.com/spacejack/carphysics2d/blob/master/public/js/Car.js",
    "camera": "https://github.com/mrdoob/Starter-Kit-Racing/blob/main/js/Camera.js",
    "quality_floor": 0.7
  },
  "research": {
    "__doc__": "v1.11.0 — a researched, CITED knowledge base the loop reads before leaping (the missing external-knowledge input that breaks the 'iterate without improving' maze). Built by a heavy web-research pass (graphics/physics/art/methodology) distilled into an actionable, sourced playbook. dev-cycle Step 3-PRE prefers a concrete playbook move (technique + parameters + source URL) over the model's first idea.",
    "playbook_path": "agent/state/research/playbook.md",
    "refresh_when_stalled": true
  },
  "leap": {
    "__doc__": "Quantum-leap cycles (evolve Steps 2-G/3-K, v1.7.0) — the fix for monotonic incrementalism (RICE structurally forbids overhauls). When the artifact plateaus BELOW bar, the next cycle is forced into LEAP mode: it overhauls the weakest dimension (step-change, not a new feature), bypassing pure RICE via a gap-to-bar bonus, with a larger size budget and an ARTIFACT-improvement gate instead of the unit-test gate. Safety: min_dimension_delta must be cleared or the leap is reverted; max_attempts_per_dimension failures escalate to HALT; cost/day caps bound spend. v1.8.0: lock_until_bar keeps leaping the SAME dimension until it clears bar (drive-to-good, not detect-and-nudge). v1.11.0: rewrite_on_stall — when recommend_rewrite() fires (incremental leaps THEMSELVES stalled), escalate from patch to a from-scratch REWRITE carrying a Reflexion negative-example memo, instead of thrashing to HALT.",
    "rewrite_on_stall": true,
    "max_lines": 1500,
    "min_dimension_delta": 0.05,
    "max_attempts_per_dimension": 2,
    "max_per_day": 2,
    "gap_weight": 30.0,
    "cost_limit_usd": 0.5,
    "lock_until_bar": true,
    "mega_leap": {
      "__doc__": "v1.9.0 — a multi-cycle RE-PLATFORM for radical jumps a normal leap can't make (replace the whole rendering pipeline, swap a subsystem). Unlocked only by a human-authored, approved mega_leap_plan.json after a leap exhausts max_attempts_per_dimension. Bigger budget, no per-cycle revert; the whole sequence reverts if the cumulative artifact delta misses min_artifact_delta_at_completion.",
      "enabled": false,
      "max_lines": 5000,
      "max_cycles": 4,
      "min_artifact_delta_at_completion": 0.15,
      "requires_human_plan_approval": true,
      "plan_file": "agent/state/evolve/mega_leap_plan.json"
    }
  },
  "goal_completion_idle": true
}