| name | user-interaction |
| description | Defines the checkpoint protocol for the elegant-researcher-skill pipeline. Use at major research transition points to summarize artifacts, show options, apply mode-specific defaults, and continue without blocking routine decisions. |
| license | MIT |
User Interaction
This skill controls how the elegant-researcher-skill pipeline involves the user.
Principles
- Never ask about routine decisions.
- Present a short decision summary, not raw files.
- Every checkpoint has a default route.
- Defaults depend on the selected mode.
- If the user says "keep going", follow the default.
- In autonomous modes, do not wait unless there is a true blocker.
Modes
| Mode | User Interaction Behavior |
|---|
quick-cycle | Minimal checkpoints; use defaults quickly |
deep-cycle | Present checkpoints, but continue with defaults when user gives no direction |
paper-ready | Ask only when full results, venue, or citation status affects claims |
theory-only | Skip experiment handoff; focus on theory and proposed experiments |
human-in-loop | Wait for user decisions at checkpoints |
Checkpoint Format
Each checkpoint response should include:
## [Checkpoint Title]
Status: [one sentence]
Mode: [mode]
Evidence level: [highest relevant level]
Key artifacts:
- [path] - [what it contains]
Summary:
- [key point 1]
- [key point 2]
- [key point 3]
Options:
A. [default option] - [route]
B. [alternative] - [route]
C. [alternative] - [route]
D. [alternative] - [route]
Default: A unless mode-specific rules say otherwise.
Do not include long raw file contents.
Checkpoint 1: Literature Map Confirmation
Trigger: literature search and literature/gap-analysis.md are complete.
Present:
- field overview,
- method evolution,
- contradictions,
- implicit assumptions,
- candidate gaps.
Options:
A. Continue with the recommended gap.
B. Deepen a specific gap.
C. Adjust literature direction.
D. Switch to theory-only framing.
Default:
- autonomous modes: A.
human-in-loop: wait for user selection.
Checkpoint 2: Hypothesis Selection
Trigger: candidate hypotheses and hypothesis-tree.md are complete.
Present:
- 2-5 hypotheses,
- recommended starting point,
- feasibility and expected information gain.
Options:
A. Start with the recommended hypothesis.
B. Select a different hypothesis or multiple hypotheses.
C. Add a user hypothesis.
D. Brainstorm again.
Default:
- autonomous modes: A.
human-in-loop: wait for user selection.
Checkpoint 3: Experiment Ready
Trigger: protocols, code, micro-validation, and handoff package are ready.
Present:
- experiment list,
- protocol locations,
- code locations,
- micro-validation result summary,
- handoff instructions,
- evidence level.
Options:
A. Use micro-validation as proof-of-concept and proceed to writing.
B. User will run full experiments; wait for results.
C. Modify or fix experiments, then re-run micro-validation.
D. Switch to theory-only or proposed-experiments path.
Default by mode:
quick-cycle: A.deep-cycle: A.paper-ready: B if user has compute and wants empirical claims; otherwise A with full claims marked pending.theory-only: D.human-in-loop: B.
This checkpoint must not silently stall autonomous runs.
Checkpoint 4: Draft Review
Trigger: first draft or research note is complete.
Present:
- artifact type,
- section status,
- figure count,
- evidence levels,
- unresolved citations,
- pending results.
Options:
A. Polish and finalize.
B. Strengthen a section.
C. Integrate new user results.
D. Rethink contribution.
Default:
- autonomous modes: A.
human-in-loop: wait for user review.
Checkpoint 5: Final Confirmation
Trigger: final artifact is ready.
Present:
- final artifact path,
- key findings,
- evidence levels,
- figure count,
- citation status,
- limitations,
- next recommended cycle.
Options:
A. Mark cycle complete.
B. Deepen a specific aspect.
C. Convert to LaTeX or venue-specific submission.
D. Start a new research question.
Default:
- autonomous modes: A.
human-in-loop: wait for user decision.
Handoff Package
When full experiments are needed, produce HANDOFF.md with:
- exact commands,
- expected runtime,
- required hardware,
- output paths,
- result schema,
- metrics to report,
- common failure fixes.
Final Response Requirements
At the end of a cycle, the agent must include:
- completion status and mode,
- key findings,
- evidence level per finding,
- output paths,
- reliable vs. preliminary claims,
- limitations,
- next step,
- citation status,
- automation status.
