chore(ai-research-workflows): v0.2.0 prep — command triggers, retreat paths, private skill#111
Open
lsetiawan wants to merge 5 commits into
Open
chore(ai-research-workflows): v0.2.0 prep — command triggers, retreat paths, private skill#111lsetiawan wants to merge 5 commits into
lsetiawan wants to merge 5 commits into
Conversation
… as private The skill exists to load templates and workflow guidance into the orchestrator agent; it is not intended to be invoked directly by users. Flipping user-invocable to false removes it from the user-facing skill index. Refs #93
…mands Every failure-capable command now documents where to go when it cannot produce a usable artifact. Maps failure modes to the next command (iterate-plan, plan, research, experiment, optimize) so users do not dead-end on first friction. Refs #93
…o every command Each command now carries its own trigger list at the top, migrated from the orchestrator's per-command guidance blocks. This makes commands discoverable and usable without routing through the orchestrator, and sets up the orchestrator trim in a follow-up commit. The /handoff section is authored fresh — no prior orchestrator block existed for it. Refs #93
…ance Each command now owns its own 'When to use' section after the previous commit. The orchestrator no longer duplicates trigger logic — it points to the command files and retains its real responsibilities: routing, workflow-state tracking, and retreat-path decisions. Net: orchestrator shrinks by roughly 80-90 lines. Sets up the structural refactor in #87. Refs #93
…ill per Tessl review Address Tessl skill review findings (35% → targeting higher): - Cut Workflow Philosophy, Integration, Common Patterns sections - Collapse verbose 'When to Use Each Step' into decision tree (commands already carry per-command triggers) - Add concrete command syntax examples with real arguments and resulting artifact paths - Add Success Criteria Automated/Manual split with executable verification command examples - Move per-phase checklists to references/checklists.md for progressive disclosure SKILL.md trimmed from ~2000 to ~620 words.
🔍 Tessl Skill Review
|
| Dimension | Score | Detail |
|---|---|---|
| conciseness | ███ 3/3 | The content is lean and efficient throughout. It uses tables, decision trees, and concrete examples without explaining concepts Claude already knows. Every section earns its place — no padding or unnecessary exposition about what workflows are or why they matter. |
| actionability | ███ 3/3 | Provides fully concrete command examples with exact syntax, file paths, and expected outputs. The decision tree gives specific guidance on which command to use in each situation. The success criteria format shows exactly how to structure automated vs manual checks with executable commands. |
| workflow clarity | ███ 3/3 | The six-phase workflow is clearly sequenced via the table and decision tree. The decision tree acts as a validation checkpoint — it explicitly gates each phase on prerequisites (e.g., 'research docs already exist?' before planning). The cross-referencing rules create traceability, and the success criteria format provides explicit validation with pass/fail recording. |
| progressive disclosure | ███ 3/3 | The skill provides a clear overview with well-signaled references to command files ('Command-level trigger details live in each command file's ## When to use section'), templates in a specific assets directory, and a separate checklists reference file. Content is appropriately split — templates are not inlined, and detailed per-command behavior is delegated to individual command files. |
Overall: This is an excellent orchestration skill that efficiently maps a six-phase workflow with concrete commands, clear sequencing, and appropriate progressive disclosure. The decision tree is particularly effective for guiding command selection, and the cross-referencing rules ensure auditability. The content is well-structured, avoids unnecessary explanation, and delegates detail to the right places.
To improve your score, point your agent at the Tessl optimization guide. Need help? Jump on our Discord.
Feedback
Report issues with this review at tesslio/skill-review, or send private feedback from your terminal with tessl feedback.
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.
Duplicate of PR #95
Summary
Prep work for v0.2.0 of the
ai-research-workflowsplugin. Pure documentation + frontmatter changes — zero behavior change.## When to usesection (migrated from the orchestrator's per-command trigger blocks, plus a freshly authored block for/handoff)./research,/experiment,/implement,/validate) each have a# Retreat Pathssection mapping failure modes to the next command.research-workflow-managementskill is nowuser-invocable: false.Closes #93. Blocks #87.
Commits
f1a9199— mark research-workflow-management skill as private36cb76e— add retreat paths to failure-capable commandsbfe3f10— add CSO 'When to use' trigger sections to every command9e7c035— trim orchestrator's per-command guidanceTest plan
## When to use/research,/experiment,/implement,/validatecontain# Retreat Pathsuser-invocable: false## When to Suggestsubsections (0 matches)plugin.jsonstill parses as valid JSON/researchand/planon a scratch topic and confirm behavior is unchangedOut of scope
Other v0.2.0 work is tracked in separate issues and will land in separate PRs:
/prototypecommand — feat(ai-research-workflows): add prototype phase #89/optimizecommand — feat(ai-research-workflows): optimize phase, plan template, and docs #92/experiment— feat(ai-research-workflows): experiment redesign — worktree isolation #88/validate+ persistent validate doc — feat(ai-research-workflows): validate redesign — empirical validation #90/iterate-planreframe — feat(ai-research-workflows): make iterate-plan a floating utility #91--profileflag + orchestrator structural refactor — feat(ai-research-workflows): core infrastructure — workflow model, profiles, retreat paths #87This PR leaves
plugin.jsonon v0.1.0.