aris-research-pipeline

Full research pipeline: Workflow 1 (idea discovery) → implementation → Workflow 2 (auto review loop). Goes from a broad research direction all the way to a submission-ready paper. Use when user says "全流程", "full pipeline", "从找idea到投稿", "end-to-end research", or wants the complete autonomous research lifecycle.

Quality

78%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/aris-research-pipeline/SKILL.md

Quality

Content

73%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured orchestration skill: the five-stage workflow is clearly sequenced with strong checkpoints, feedback loops, and stop conditions, and the guidance is largely actionable. The main costs are triple-explained AUTO_PROCEED semantics wasting tokens and a dangling template reference with no bundle file behind it.

Suggestions

Explain AUTO_PROCEED once in the Constants section and have Gate 1 / Key Rules reference it ("per AUTO_PROCEED above") instead of restating both branches three times.

Either ship templates/RESEARCH_BRIEF_TEMPLATE.md (and the final-report template as a separate file) or remove the reference, so every referenced path exists.

Tighten Stage 2 into a concrete checklist (files to create, argparse/seeds/logging requirements as verifiable items) to close the gap between it and the concrete Stage 3-4 instructions.

Dimension	Reasoning	Score
Conciseness	The body is operational and teaches nothing Claude already knows, but the AUTO_PROCEED semantics are explained three separate times ("AUTO_PROCEED = true — When `true`, Gate 1 auto-selects…", "If AUTO_PROCEED=false: Wait for user confirmation…", "Human checkpoint after Stage 1 is controlled by AUTO_PROCEED. When `false`, do not proceed…"), and the "Sweet spot" anecdote adds no instruction — matching 'mostly efficient but could be tightened' rather than the minor-trimming of anchor 4.	3 / 5
Actionability	Concrete sub-skill invocations ("/aris-idea-discovery \"$ARGUMENTS\""), a rendered Gate 1 dialog template, a four-item code-review checklist, and a full final-report markdown template give mostly executable guidance; minor gaps remain in Stage 2, where directives like "Extend pilot code to full scale (multi-seed, full dataset, proper baselines)" are directional rather than copy-paste, keeping it below anchor 5.	4 / 5
Workflow Clarity	Five clearly sequenced stages with an explicit human checkpoint enumerating every user response (approve / pick different / request changes / reject all / stop), a pre-deploy self-review checklist, experiment-start verification, and an auto-review feedback loop with an explicit stop condition ("repeat until score ≥ 6/10 or 4 rounds reached") plus fail-gracefully rules — matching the anchor for explicit validation steps, feedback loops, and checklists.	5 / 5
Progressive Disclosure	Well-sectioned overview with one-level-deep, clearly signaled pointers to sub-skills, and no bundle files to reorganize; however, the referenced "templates/RESEARCH_BRIEF_TEMPLATE.md" does not exist in the bundle, and the inline Gate 1 dialog and final-report templates are content that could live in separate files — good structure with minor organization gaps (anchor 4), not the clean split of anchor 5.	4 / 5
	Total	16 / 20 Passed

Description

83%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: it states a clear multi-stage capability with an explicit outcome and provides a well-chosen, bilingual set of quoted trigger phrases. The main gaps are un-enumerated stage internals and slight overlap risk with the sub-skills it orchestrates.

Suggestions

Enumerate one or two concrete actions per stage (e.g., 'generates ranked pilot-tested ideas, deploys GPU experiments, runs adversarial review loops') to lift specificity from stage names to concrete capabilities.

Add a couple of common English variants users would naturally say, such as "research pipeline" or "from idea to paper", to broaden trigger coverage.

Dimension	Reasoning	Score
Specificity	"Workflow 1 (idea discovery) → implementation → Workflow 2 (auto review loop)" and "Goes from a broad research direction all the way to a submission-ready paper" name several concrete pipeline stages plus the outcome, but the actions within each stage are not enumerated — matching 'several specific actions; minor gaps' rather than the comprehensive coverage of a 5.	4 / 5
Completeness	It explicitly answers both what ("Full research pipeline… Goes from a broad research direction all the way to a submission-ready paper") and when ("Use when user says…") with concrete quoted trigger phrases, matching the anchor for clearly and explicitly answering both; the 'when' is more explicit than the anchor-4 example allows.	5 / 5
Trigger Term Quality	Explicit natural phrases in both languages ("全流程", "从找idea到投稿", "full pipeline", "end-to-end research") plus a behavioral trigger ("wants the complete autonomous research lifecycle") give good coverage, though common variants like "research pipeline" or "find ideas and write the paper" are missing, so it falls short of the synonym-complete anchor 5.	4 / 5
Distinctiveness Conflict Risk	The end-to-end lifecycle triggers occupy a clear niche distinct from generic skills, but terms like "end-to-end research" and "full pipeline" could plausibly match its own chained sub-skills (/aris-idea-discovery, /aris-auto-review-loop), giving minor overlap risk — anchor 4 rather than 5.	4 / 5
	Total	17 / 20 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: OpenLAIR/dr-claw
Path: skills/aris-research-pipeline/SKILL.md
Commit: d51b64e

Reviewed: 6 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.