Incorporate Learnings

Incorporate Learnings

Take investigated issues and integrate the lessons learned into the LearningAgent's knowledge.

Arguments

$ARGUMENTS is the path to the session log folder (e.g., .deepwork/tmp/agent_sessions/<session_id>/<agent_id>/).

Context

Agent used: !cat $ARGUMENTS/agent_used 2>/dev/null || echo "unknown"

If agent_used is "unknown", stop and report an error.

Investigated Issues you need to Process: !grep -l 'status: investigated' $ARGUMENTS/*.issue.yml 2>/dev/null || echo "(none)"

Additional incorporation guidelines: !learning_agents/scripts/cat_agent_guideline.sh $ARGUMENTS learning_from_issues

Procedure

If no investigated issues are listed above, skip to Step 3 (still update tracking files).

Step 1: Incorporate Each Issue

For each investigated issue, read its issue_description and investigation_report, then choose the best incorporation strategy in this priority order:

A. Amend existing content (prefer this)

If a closely related file already exists in the agent's knowledge base, edit it rather than creating a new one.

B. Update core-knowledge.md

For universal one-liners — something fundamental expressible in 1-2 sentences.

C. Add a new topic in topics/

For a conceptual area needing 1+ paragraphs of reference material. Use this frontmatter:

---
name: <Topic Name>
keywords:
  - <relevant keyword>
last_updated: <YYYY-MM-DD>
---

D. Add a new learning in learnings/

For cases where the narrative context of how the issue unfolded is needed. Use this frontmatter:

---
name: <Descriptive name>
last_updated: <YYYY-MM-DD>
summarized_result: |
  <1-3 sentence summary of the key takeaway>
---

## Context
## Investigation
## Resolution
## Key Takeaway

If you add a learning, also consider adding a brief note to a related Topic referencing it.

E. Do nothing

If the issue is unlikely to recur or would be hard to prevent, skip it.

Step 2: Update Issue Status

For each incorporated issue, change status: investigated to status: learned.

Step 3: Update Session Tracking

Always run this step, even if no issues were incorporated.

[ -f $ARGUMENTS/needs_learning_as_of_timestamp ] && rm $ARGUMENTS/needs_learning_as_of_timestamp
date -u +"%Y-%m-%dT%H:%M:%SZ" > $ARGUMENTS/learning_last_performed_timestamp

Step 4: Summary

## Incorporation Summary

- **Issues processed**: <count>
- <issue-filename> → <action taken: updated core-knowledge | created topic <name> | created learning <name> | amended <filename> | skipped>

Guardrails

• Do NOT create overly broad or vague learnings — be specific and actionable
• Do NOT duplicate existing knowledge — check the auto-included lists above
• Do NOT remove existing content unless directly contradicted by new evidence
• Keep core-knowledge.md concise — move detailed content to topics or learnings
• Use today's date for last_updated fields
• Always run Step 3 even if no issues were incorporated