Menu
Invoke for investigations.json and current_investigation.json format reference. Defines the schema for claim investigations and query bundles.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | write-investigation |
| description | Invoke for investigations.json and current_investigation.json format reference. Defines the schema for claim investigations and query bundles. |
Two files share the same investigation structure:
/workspace/run/investigations.json — cumulative record of all investigations. Root object with investigations array and active_investigation_id string./workspace/run/current_investigation.json — the investigation currently being executed. Single investigation object (same structure as one element of the investigations array). Overwritten each cycle.| Field | Type | Required | Description |
|---|---|---|---|
id | string | yes | Investigation identifier (I1, I2, ...) |
name | string | yes | Short label for the investigation |
phenomenon | string | yes | Full sentence describing the observed phenomenon being investigated |
adjudication_goal | string | yes | What this investigation aims to determine — stated as a question or judgment to be made |
pattern_ids | array of strings | yes | Pattern IDs this investigation addresses |
status | string | yes | One of: active, paused, resolved, retired, saturated |
cycle_id | integer | yes | Current cycle number for this investigation |
claims | array | yes | Claims being evaluated (see below) |
query_bundle | array | yes | Queries to execute (see below) |
active: currently being investigatedpaused: temporarily set aside, may resumeresolved: phenomenon explained, evidence sufficientretired: abandoned — artifact, low value, or merged with another investigationsaturated: diminishing returns, not fully resolved but not worth continuingclaims array)| Field | Type | Required | Description |
|---|---|---|---|
id | string | yes | Claim identifier (e.g. I1-claim1) |
text | string | yes | The claim statement |
role | string | yes | One of: main, alternative_explanation, artifact_check, robustness |
status | string | yes | One of: pending, supported, weakened, refuted, inconclusive |
confidence | number or null | yes | Judge's calibrated confidence (0-1), null if not yet evaluated |
supporting_evidence | array | yes | Experiment IDs that support this claim |
counter_evidence | array | yes | Experiment IDs that weaken or refute this claim |
main: the primary hypothesis about the phenomenonalternative_explanation: a competing explanation for the same phenomenonartifact_check: tests whether the phenomenon is an artifact of data processing, sampling, or annotationrobustness: tests whether the main finding holds under different conditions or parametersquery_bundle array)| Field | Type | Required | Description |
|---|---|---|---|
id | string | yes | Query identifier (e.g. I1-q1) |
claim_id | string | yes | Which claim this query tests |
text | string | yes | The specific analysis to perform |
role | string | yes | One of: observational, diagnostic, artifact_check |
expected_effect | string | yes | What the result would mean for the claim — how it discriminates between explanations |
result_ref | string or null | yes | Experiment ID that produced the result, null if not yet run |
depends_on | array of strings | yes | Query IDs that must complete before this one (empty array if independent) |
status | string | yes | One of: pending, running, done, failed, skipped |
| Field | Type | Required | Description |
|---|---|---|---|
active_investigation_id | string or null | yes | ID of the currently active investigation |
Invoke after experiment-runner completes an investigation's query bundle. Evaluates all claims in current_investigation.json and produces an investigation-level adjudication. Writes claim_scores.json.
Invoke after investigation-decomposition. Runs all queries in an investigation bundle using shared data preparation. Writes code and results.
Invoke after main agent writes L2 assessment. Reads epistemic state (L1/L2), patterns, and investigations to recommend next action. Writes strategy_recommendation.json.
Gene set enrichment analysis (GSEA, Enrichr, over-representation). Invoke when query mentions enrichment, pathway analysis, GO analysis, GSEA, or gene set.
Invoke before running experiments for an investigation. Reads current_investigation.json and task_packet.json, writes current_investigation_requirements.json with shared data preparation and per-query requirements.
Gene ID conversion and annotation. Invoke when you need to map between any gene identifier formats.