autoresearch-goal
Durable professor-critic research workflow over Codex goal mode without reviving deprecated omx autoresearch
Best for Codex goal-mode workflowsWorks with GitHub
⌘source
- author
@Yeachan-Heo- language
- TypeScript
★28kstars
2.2kforks
May 6, 2026
✦overview.md
Key Features
- ·Professor-critic rubric system
- ·Goal-mode bound research
- ·Iterative research with verdicts
- ·Outputs mission, rubric, ledger, completion
Use Cases
- →Structured research with automatedcriteria evaluation
- →Iterative investigation with pass/fail checkpoints
Best for
- ✓Codex goal-mode workflows
- ✓Structured research tasks
Not ideal for
- !Ad-hoc research without criteria
- !One-off scripts
FAQs
name
autoresearch-goal
description
Durable professor-critic research workflow over Codex goal mode without reviving deprecated omx autoresearch
Autoresearch Goal
Use this workflow when a research mission should be bound to Codex goal-mode focus while OMX remains the durable state owner.
Boundary
- Do not use or revive the deprecated
omx autoresearchdirect launch surface. - Do not claim shell commands mutate hidden Codex
/goalstate. - Do not edit upstream
../../codexor add dependencies. - Use
get_goal,create_goal, andupdate_goal({status: "complete"})only through the active Codex thread when those tools are available.
Artifacts
omx autoresearch-goal writes:
.omx/goals/autoresearch/<slug>/mission.json.omx/goals/autoresearch/<slug>/rubric.md.omx/goals/autoresearch/<slug>/ledger.jsonl.omx/goals/autoresearch/<slug>/completion.json
Flow
- Create the mission and professor-critic rubric:
omx autoresearch-goal create --topic "..." --rubric "..." --critic-command "..." - Emit the model-facing handoff:
omx autoresearch-goal handoff --slug <slug> - In the active Codex thread, call
get_goal; callcreate_goalonly if no active goal exists and the printed payload is the intended objective.
$install
1-click copynpx skills add Yeachan-Heo/oh-my-codex --skill autoresearch-goalSafety assessment
★
Clarity score
How clear and easy to understand the SKILL.md instructions are, rated from 1 to 5.
4/ 5
very goodClear and well structured, with only minor parts that might need a second read.
◎
Actionability score
How directly an agent can act on the SKILL.md instructions, rated from 1 to 5.
3/ 5
mediumPartially actionable with several concrete steps, but still missing important details.