Skip to content

CONFIGURATION.md

Latest commit

 

History

History
98 lines (80 loc) · 4.6 KB

CONFIGURATION.md

File metadata and controls

98 lines (80 loc) · 4.6 KB
Raw

Configuration

Part of the CLK documentation. Cost guardrails, customization points, and dry-run mode. Provider/.env setup lives in QUICKSTART, WEBUI, and KICKOFF.

Cost guardrails

Title-bar dollar cost is computed from the per-provider table in clk_harness/pricing.py:

Provider Default $/1k in Default $/1k out
claude (sonnet-4-5) $0.003 $0.015
claude (haiku-latest) $0.0008 $0.004
claude (opus-latest) $0.015 $0.075
codex (gpt-4o) $0.0025 $0.010
codex (gpt-4o-mini) $0.00015 $0.0006
codex (o1) $0.015 $0.060
gemini (1.5-pro) $0.00125 $0.005
gemini (1.5-flash) $0.000075 $0.0003
pi $0.003 $0.015 (blended default; override per route)
ollama / shell $0.00 $0.00

Override per project by adding to .clk/config/providers.json:

"providers": {
  "pi": {
    "type": "pi",
    "pricing": { "input_per_1k": 0.002, "output_per_1k": 0.008 }
  }
}

Or per model:

"pricing_by_model": { "openrouter/free": { "input_per_1k": 0.0, "output_per_1k": 0.0 } }

/status prints the per-provider breakdown so you can see which provider is eating the budget. Updated lazily from the same numbers the title bar shows.

Robustness-loop multipliers

The robustness loops (see Robustness loops) trade tokens for quality. Use this table to pick a regime:

Knob Worst-case multiplier per affected dispatch Recommended starting point
robustness.auto_consensus off → ×1; on_careful → ×(N+1) on careful stages only; always → ×(N+1) on every dispatch (where N = consensus.max_samples, default 6) on_careful (default)
robustness.auto_refine off → ×1; careful_only → ×(1 + 1 worker revision + 1 critic) on careful stages; all → that on every producing stage all (default — every producing stage refines)
robustness.max_quality_retries At most this many extra dispatches when a response fails the quality check; 0 disables 4 (default)
robustness.refine_max_rounds Cap on critic↔worker round-trips inside a refine loop 10 (default)
robustness.debate off → ×1; careful_only → an adversarial panel (one critic per lens) on careful stages; all → on every producing stage careful_only (default)
robustness.debate_lenses Adversarial lenses (one parallel critic each) — more lenses = more critic dispatches per round [correctness, security, simplicity]
robustness.debate_max_rounds Cap on debate rounds (panel critique + worker revision) per stage 2 (default)
robustness.max_qa_depth Cap on inter-agent Q&A chain depth (each peer answer can ask one peer) 3 (default)
robustness.plateau_window How many no-improvement Ralph/autoresearch iterations before escalation 3 (default)
robustness.plateau_action off disables adaptive loop termination entirely escalate_then_reframe

Cost-minimal regime (closest to legacy CLK behavior, no extra tokens):

"robustness": {
  "auto_consensus": "off",
  "auto_refine": "off",
  "max_quality_retries": 0,
  "plateau_action": "off"
}

Cost-maximal "lean into the loop" regime (every dispatch fans out, critic gates every careful stage, plateau detection on, Q&A protocol fully open):

"robustness": {
  "auto_consensus": "always",
  "auto_refine": "all",
  "max_quality_retries": 3,
  "refine_max_rounds": 4,
  "plateau_action": "escalate_then_reframe"
}

Customization

  • Edit prompts in .clk/prompts/ to change agent behavior.
  • Edit .clk/config/agents.json to bind specific agents to specific providers (e.g. engineer -> claude, researcher -> ollama).
  • Edit .clk/config/workflows/*.yaml to add new stages or new workflows. Reference any new workflow with clk run --workflow NAME.
  • clk configure --set key=value updates .clk/config/clk.config.json.

Dry-run mode

Every loop and workflow command accepts --dry-run. Providers honor it and skip side effects. Use it to preview prompt rendering and stage ordering without writing files or committing.