task-outcome-lookup

name	task-outcome-lookup
description	Deterministic read-only lookup of historical task outcomes for Agent Factory routing. Uses task_outcome_ledger_read.py only and never invokes an LLM.

Task Outcome Lookup

Use this skill after task-evaluate and rank-task-difficulty. It reads historical task outcomes and returns routing evidence for factory.

This skill is read-only.

Inputs

<ledger>
.kamino/evals/tasks/task-outcome-ledger.jsonl
</ledger>

<task_evaluation_json>
path/to/task-evaluation.json
</task_evaluation_json>

<difficulty_json>
path/to/difficulty-placement.json
</difficulty_json>

Optional filters:

<task_type>writing</task_type>
<difficulty_band>0.5</difficulty_band>
<agent>.kamino/agents/library/writing/article-review-agent.md</agent>
<model>sonnet</model>
<effort>medium</effort>

Uses

• Script only: .kamino/evals/scripts/task_outcome_ledger_read.py

Run the script only through uv run:

uv run .kamino/evals/scripts/task_outcome_ledger_read.py \
  --ledger ".kamino/evals/tasks/task-outcome-ledger.jsonl" \
  --task-eval "<task-eval.json>" \
  --difficulty "<difficulty.json>" \
  --format json

Output

Strict JSON containing:

• similar historical tasks
• successful agent/model/effort combinations
• failed agent/model/effort combinations
• risk notes derived only from ledger records
• filters applied

Rules

1. Do not invoke any LLM agent.
2. Do not write files.
3. Do not mutate the ledger.
4. Do not infer historical outcomes that are not in the ledger.
5. Do not treat partial historical completion as success.
6. Do not invoke AutoResearch.
7. Run Python only through uv run.

Failure Conditions

Fail clearly if:

1. The ledger contains malformed or schema-invalid records (a missing or empty ledger is a cold start: the lookup succeeds with no historical outcomes).
2. The task evaluation JSON is missing or schema-invalid.
3. The difficulty placement JSON is missing or schema-invalid.
4. A filter is malformed.
5. The script exits non-zero.