AI Task Fit

ENTRY: mf-tool-002 · SECTION: tools · agent audience

Should AI touch this task? Rules + decision tree + 12 examples. Complements mf-tool-001 (software/automation), not replace it.

Free worked example

{
  "task": "Draft reply to angry customer email about late delivery",
  "inputs": {
    "repeatability": "occasional",
    "judgment": "review",
    "stakes": "customer",
    "input_shape": "messy",
    "output_type": "draft",
    "volume": "low",
    "data_known": "yes"
  },
  "outcome": "llm_assist",
  "why": "Messy thread, customer stakes. Human edits before send.",
  "operator_lens": "Use mf-tpl-001 framework after draft; never auto-send."
}

Full pack (x402)

rules.json: 7 outcomes, decision_tree map, priority rules
examples.json: 12 historical small-business scenarios
OUTCOMES.md + AGENT.md + engine.js

FULL (x402): tools/ai-task-fit/full/
PRIMARY: full/rules.json (gated)
PRICE: $0.05 · USDC on Base
RELATED: mf-tool-001 · mf-tpl-001