Supplement commands with autoresearch-inspired patterns - #32
Merged
Conversation
Add guard commands, investigation techniques, scenario expansion, and adversarial debate to four existing commands — no new files or structural changes. - autoloop: guard safety-net + escalation after repeated failures - tri-debug: investigation techniques + domain-specific checklists - tri-test-gen: scenario expansion techniques in generation prompt - deep-research: adversarial debate cycle for close hypotheses
autoloop: - Add baseline guard check in Step 2 (fail-fast if invariant broken) - Add timeout to guard execution (120s default) - Require guard commands to be side-effect free - Define escalation trigger concretely (same primary modified file) - Remove subjective "sacred" wording deep-research: - Move adversarial debate after evidence matrix (was Step 7.5, now 8.5) - Require matrix as input to judge evaluation - Relax "mutually exclusive" to "competing" hypotheses - Composites replace parent hypotheses and must be rescored - Acknowledge blind judge is a heuristic nudge, not true blinding - Add concrete skip threshold (2+ fewer inconsistencies) tri-debug: - Extract domain checklists to commands/references/debug-checklists.md - Fix contradictory "consensus symptoms" rule to distinguish symptom-agreement from full disagreement - Expand vague auth checklist item tri-test-gen: - Move scenario expansion to orchestrator guidance (Step 2.5) - Only inject applicable techniques per target, not full table - Add integration-test caveat for interruption/stale-data - Respect existing repo test conventions over default priority
autoloop: - Document intentional guard skip on regressions (not a gap) - Integrate per-file failure counter into Step 7 REGRESSED flow deep-research: - Reorder judge criteria: evidence-fit primary, falsifiability tiebreaker tri-debug: - Fix cross-reference path to be relative to commands/ directory tri-test-gen: - Fix contradictory parenthetical on Step 2.5 injection scope - Replace phantom "error paths" with "what-if" in priority list debug-checklists: - Add sub-heading for performance symptom lookup table
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Guard: npm test) that must pass alongside the metric, plus escalation logic that skips and pivots after 3 failures on the same targetInspired by patterns in uditgoenka/autoresearch, adapted to fit existing devkit command structure. No new files or commands — all changes supplement existing workflows.
Test plan
/autoloopwith a guard command and verify guard failures trigger revert/tri-debugon an API bug and confirm domain checklist is included in agent prompt/tri-test-genand verify scenario expansion techniques appear in generated tests/deep-researchon a question with close hypotheses and confirm adversarial debate triggers