Document local decision intervention research

[Tyler-R-Kendrick]

What changed

• Captures all 25 academic works and eight implementation/documentation sources recovered from the shared ChatGPT guidance.
• Adds a primary-source-backed design synthesis and explicit Adapted/Adjacent boundaries.
• Records the Auto-Antislop and Antidoom revisions and license constraints.

Why

The repository needs durable citation lineage and falsifiable local-decision levers before implementation or experiment claims.

Validation

• pytest -q tests/test_autoresearch/test_harness.py
• ruff check tests/test_autoresearch/test_harness.py
• python -m scripts.repo_policy

No train, eval, benchmark, or checkpoint run was performed.

Summary by CodeRabbit

• Documentation
  • Added implementation guidance for local decision interventions, including evidence tracking, decision-event formats, training approaches, evaluation metrics, and safeguards.
  • Documented the exact-state local decision preference approach, related research, implementation status, and planned experiment coverage.
• Research Resources
  • Added a curated manifest of 33 academic and implementation sources, including summaries and relevance notes.
• Tests
  • Added validation ensuring the research manifest is complete, unique, consistently annotated, and correctly categorized.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
slm-training	Ready	Preview, Comment	Jul 16, 2026 8:23pm

[coderabbitai]

Note

Currently processing new changes in this PR. This may take a few minutes, please wait...

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: f274bacd-4273-4c75-a718-c4131083099c

📥 Commits

Reviewing files that changed from the base of the PR and between 63856c0 and 88318ab.

📒 Files selected for processing (4)

• docs/design/local-decision-interventions.md
• docs/design/research-lineage.md
• src/slm_training/resources/autoresearch/local-decision-sources.json
• tests/test_autoresearch/test_harness.py

 ____________________________________________
< Double tap everything. Kill the bugs dead. >
 --------------------------------------------
  \
   \   \
        \ /\
        ( )
      .( o ).

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch agent/local-decision-research

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Caution

Failed to replace (edit) comment. This is likely due to insufficient permissions or the comment being deleted.

Error details

putComment timed out