github uditgoenka/autoresearch v2.0.1
v2.0.1 — Universal --chain Support

latest releases: v2.1.4, v2.1.3, v2.1.2...
one month ago

What's New

Universal --chain Support

All 10 autoresearch subcommands now support --chain <targets> for sequential pipeline execution. Previously only predict, probe, and reason had this capability.

Example pipelines:

# Debug → Reason about fix approach → Apply fixes
/autoresearch:debug --chain reason,fix

# Security audit → Fix vulnerabilities → Ship
/autoresearch:security --chain fix,ship

# Explore scenarios → Debug edge cases → Fix issues
/autoresearch:scenario --chain debug,fix

# Learn codebase → Plan improvements
/autoresearch:learn --chain plan

Each workflow now includes a Chain Conversion section documenting exactly how its output maps to each of 9 downstream targets, plus a Multi-Chain Execution section for sequential pipelines.

Space-After-Comma Tolerance

--chain reason, fix now works the same as --chain reason,fix. The parser splits on comma and trims each token.

Probe Detection Fix

/autoresearch:probe was not triggering in some installations because:

  • probe.md command file was missing from claude-plugin/
  • probe-workflow.md reference was missing from claude-plugin/
  • Full probe section was missing from claude-plugin/ SKILL.md

All synced. Probe now works across all 4 platforms.

Chain Support Matrix (v2.0.1)

Command --chain Targets
/autoresearch:debug NEW fix, security, scenario, predict, plan, learn, reason, ship, probe
/autoresearch:fix NEW debug, security, scenario, predict, plan, learn, reason, ship, probe
/autoresearch:learn NEW debug, fix, security, scenario, predict, plan, reason, ship, probe
/autoresearch:plan NEW debug, fix, security, scenario, predict, learn, reason, ship, probe
/autoresearch:scenario NEW debug, fix, security, predict, plan, learn, reason, ship, probe
/autoresearch:security NEW debug, fix, scenario, predict, plan, learn, reason, ship, probe
/autoresearch:ship NEW debug, fix, security, scenario, predict, plan, learn, reason, probe
/autoresearch:predict v2.0.0 debug, security, fix, ship, scenario
/autoresearch:probe v2.0.0 plan, predict, debug, fix, scenario, reason, ship, learn
/autoresearch:reason v2.0.0 debug, plan, fix, security, scenario, predict, ship, learn

Backward Compatibility

  • --fix shortcut in debug/security still works (equivalent to --chain fix)
  • All existing --chain behavior in predict, probe, reason unchanged
  • No breaking changes from v2.0.0

Platforms

All changes synced across:

  • Claude Code (.claude/)
  • OpenCode (.opencode/)
  • OpenAI Codex (.agents/)
  • Claude Plugin (claude-plugin/)

Stats

  • 59 files changed
  • 3,840 lines added
  • 67 lines removed

Full Changelog: v2.0.0...v2.0.1

Don't miss a new autoresearch release

NewReleases is sending notifications on new releases.