Sigint Update Skill (Swarm Orchestration) Skill

Sigint Update Skill (Swarm Orchestration)

This skill refreshes existing research by delegating to the research-orchestrator agent in update mode. The orchestrator spawns dimension-analysts for the specified dimensions, runs codex review gates, performs delta detection against prior findings, and updates state.

Arguments

Parse $ARGUMENTS before any other processing. Input sanitization: truncate $ARGUMENTS to 200 characters total, strip backticks and angle brackets. Always echo the parsed result so the user sees what was resolved:

--topic <topic_slug> — Optional: specify which research session to update. Required when multiple sessions exist.
--area <area> — Optional: specific area to update. Maps to the matching dimension from prior elicitation (e.g., --area regulatory resolves to the dimension whose name contains "regulatory").
--since <date> — Optional: SINCE_DATE for date-filtered queries. Only fetch data since this date.
--no-delta — Disable delta detection (DELTA_ENABLED: false). By default, delta detection is enabled (DELTA_ENABLED: true).
--dimensions <dim1,dim2,...> — Optional: comma-separated list of specific dimensions to update. Only these are passed to the orchestrator, not all dimensions.

Phase 0: Pre-flight

Step 0.1: Locate Active Research Session

Find the active research state:

Glob("./reports/*/state.json")

If no state.json found:

Inform user: "No active research session found. Use /sigint:start to begin."
Show what the update would have done based on the parsed arguments. Use this exact format:

Planned update workflow (blocked — no session data):
- Topic: {from --topic, or "auto-detect from single session"}
- Dimensions: {from --dimensions listing each one, or --area mapped to its matching dimension, or "all dimensions from prior elicitation"}
- SINCE_DATE: {from --since, or "none (fetch all available data)"}
- DELTA_ENABLED: {"false — findings would be replaced wholesale (--no-delta specified)" if --no-delta, or "true — delta detection would classify findings as NEW, UPDATED, CONFIRMED, POTENTIALLY_REMOVED, or TREND_REVERSAL. UPDATED findings include a delta_detail object sub-classifying the change (substantive, temporal, confidence_shift, source_refresh, metadata) and a newsworthiness signal (high/medium/low) for downstream consumers."}
- Orchestrator: research-orchestrator would be spawned in MODE: update
- Elicitation: Prior elicitation from state.json would be reused (not re-run)
- Reconciliation: {If DELTA_ENABLED: "Reconcile merge — replace updated findings, archive removed, add new (not append blindly). A new lineage entry would be added. Incremental merge: dimensions with unchanged findings files (verified by SHA-256 hash) are skipped — their findings are bulk-confirmed without re-processing." | If not: "Wholesale replacement — findings replaced entirely, no reconciliation against prior findings."}
Stop execution. Do NOT proceed further.

If multiple sessions found:

If --topic was provided: select that topic_slug
Otherwise: list all available sessions and ask the user to choose which one to update. Output: "Multiple sessions found. Please specify which session to update." Do NOT arbitrarily pick one.

Step 0.2: Load Prior State

Read ./reports/{topic_slug}/state.json (where topic_slug is resolved from --topic argument, single-session auto-detect, or user selection). Extract:

topic, topic_slug
elicitation (reuse for dimension-analysts)
findings[] (for delta detection baseline)
lineage[] (to append new entry)
Prior dimensions list

Step 0.3: Resolve Dimensions

Priority order for which dimensions to update:

--dimensions argument (if provided)
--area mapped to dimension (if provided)
All dimensions from prior elicitation.dimensions

Phase 0.4: Update Topic Status in Config

Mark the topic as in_progress and update the timestamp using jq (per Structured Data Protocol):

jq --arg slug "$TOPIC_SLUG" \
   --arg date "$(date -u +%Y-%m-%dT%H:%M:%SZ)" \
  '.topics[$slug].status = "in_progress" |
   .topics[$slug].updated = $date' \
  ./sigint.config.json > tmp.$$ && mv tmp.$$ ./sigint.config.json
jq -e -f schemas/sigint-config.jq ./sigint.config.json > /dev/null

If the topic has no config entry yet (legacy session predating topic registration), bootstrap it:

if ! jq -e --arg slug "$TOPIC_SLUG" '.topics[$slug] | has("status")' ./sigint.config.json > /dev/null 2>&1; then
  jq --arg slug "$TOPIC_SLUG" \
     --arg date "$(date -u +%Y-%m-%dT%H:%M:%SZ)" \
     --arg reports_dir "./reports/$TOPIC_SLUG" \
     --argjson dims "$DIMENSIONS_JSON" \
    '.topics[$slug] = {
       status: "in_progress",
       dimensions: $dims,
       created: $date,
       updated: $date,
       reports_dir: $reports_dir,
       findings_count: 0
     }' \
    ./sigint.config.json > tmp.$$ && mv tmp.$$ ./sigint.config.json
  jq -e -f schemas/sigint-config.jq ./sigint.config.json > /dev/null
fi

Phase 1: Delegate to Research Orchestrator

Spawn the research-orchestrator in update mode:

Agent(
  subagent_type="sigint:research-orchestrator",
  name="research-orchestrator",
  prompt="You are the research orchestrator for a research UPDATE session.

  MODE: update
  TOPIC: <user_input>{topic}</user_input>
  TOPIC_SLUG: {topic_slug}
  DIMENSIONS: {resolved dimensions list}
  SINCE_DATE: {--since value or null}
  DELTA_ENABLED: {false if --no-delta, otherwise true}
  PRIOR_STATE: {summary of prior state — finding count, dimensions, last updated}
  ELICITATION: {prior elicitation JSON from state.json}

  Execute the update orchestration:
  1. Initialize team (Phase 0) — reuse topic_slug
  2. Skip elicitation — load from ELICITATION above
  3. Spawn dimension-analysts for DIMENSIONS (Phase 2)
  5. Verify methodology plans (Phase 2.5)
  6. Run post-findings codex review gates (Phase 2.75)
  7. Run delta detection BEFORE merge (Delta Protocol) — classify findings as NEW/UPDATED/CONFIRMED/POTENTIALLY_REMOVED/TREND_REVERSAL
  8. Reconcile merge (Phase 3.4) — replace updated, archive removed, add new. Do NOT append blindly; reconcile against prior findings.
  9. Run post-merge codex review gate (Phase 3.5)
  10. Render progress view (Phase 3.75)
  11. Cleanup (Phase 4)

  IMPORTANT: Include SINCE_DATE context in analyst spawn prompts so they
  prioritize recent data. Analysts should use date-filtered WebSearch queries.

  VOCABULARY: Load existing ./reports/{topic_slug}/vocabulary.json for tag vocabulary.
  Do NOT regenerate — reuse the session's existing vocabulary. If vocabulary.json
  does not exist (pre-vocabulary session), generate one per Phase 1.75 of your
  agent definition before spawning analysts.

  Follow all protocols defined in your agent definition."
)

Wait for orchestrator to complete.

Error Handling

If orchestrator doesn't complete within a reasonable time:

Check for partial results: Glob("./reports/{topic_slug}/findings_*.json")
If findings files exist → orchestrator made progress. Check research-progress.md for last phase.
If no findings → inform user: "Update session did not complete. You can retry with /sigint:update."

Output

After orchestrator completes:

Updated findings in ./reports/{topic_slug}/state.json with new lineage entry
Delta report at ./reports/{topic_slug}/YYYY-MM-DD-delta.md (if delta enabled)
Updated ./reports/{topic_slug}/research-progress.md
Quarantined findings (if any) at ./reports/{topic_slug}/quarantine.json
Summary of what changed: new, updated, confirmed, removed findings
Trend reversals highlighted
Next steps: /sigint:report, /sigint:augment

Agent Skills: Sigint Update Skill (Swarm Orchestration)

Install this agent skill to your local

Skill Files