sdd-research

Trigger: SDD research, external evidence, source-backed research. Produce auditable evidence for a selected research lane.

Quality

65%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./internal/assets/skills/sdd-research/SKILL.md

Quality

Content

67%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.

The body is a tightly written operational spec with good sequencing, decision gates, and a hybrid-store failure-recovery loop, though much concrete execution detail is deferred to external shared references. It balances conciseness and structure well but could be more directly actionable.

Suggestions

Inline the essential store-contract mechanics (or a one-line summary) so step 4 is actionable without opening the external reference.

Make the contradiction/freshness handling in step 3 an explicit checkpoint with a pass/block decision rather than a recording instruction.

Consolidate all external references into the References section and confirm the '../_shared/' paths resolve within the bundle.

Dimension	Reasoning	Score
Conciseness	The body is lean and directive with no padding or explanations of concepts Claude already knows; every section earns its place. It is not 5 because some jargon-dense phrasing (e.g. 'canonical desired content', capability grant strings) could be tightened for clarity.	4 / 5
Actionability	Steps name concrete artifacts and contracts ('Persist gentle-ai.sdd-research/v1', 'active store contract'), but the actual execution mechanism is deferred to external shared references rather than specified inline. It is not 4 because key operational details (how to persist, what the store contract entails) are not given directly.	3 / 5
Workflow Clarity	Execution Steps 1-5 are clearly sequenced with explicit validation (verify grants, stop on denial) and a feedback loop in step 5 for one-sided hybrid-store failure. It is not 5 because some checkpoints (e.g. how contradictions/freshness in step 3 gate progress) are implicit rather than explicit checklists.	4 / 5
Progressive Disclosure	Clear sectioned overview with well-signaled, one-level-deep references to '../_shared/research-lifecycle.md', 'sdd-phase-common.md', and 'persistence-contract.md'. It is not 5 because the referenced shared files live outside the skill bundle and cannot be verified, and references are split between the Hard Rules and References sections.	4 / 5
	Total	15 / 20 Passed

Description

62%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.

The description is concise and covers both what and when with an explicit trigger clause, but the action set is thin and the trigger terms lean on framework jargon rather than natural user phrasing. It is a competent, slightly above-midline description with room for more concrete actions and synonyms.

Suggestions

Add 1-2 more concrete actions (e.g. 'validate source credibility', 'map claims to source IDs') to lift specificity toward 5.

Expand the trigger clause with natural synonyms users might say (e.g. 'external research', 'evidence gathering', 'cite sources') rather than only 'SDD research'.

Sharpen the 'when' with concrete trigger phrases to move completeness from 4 to 5.

Dimension	Reasoning	Score
Specificity	Names the domain ('SDD research, external evidence, source-backed research') and one concrete action ('Produce auditable evidence for a selected research lane'), but coverage is not comprehensive. It does not reach 4 because only a single high-level action is listed rather than several specific ones.	3 / 5
Completeness	It states both what ('Produce auditable evidence for a selected research lane') and when (an explicit 'Trigger:' clause), satisfying the explicit-trigger requirement that avoids the cap of 3. It stays at 4 rather than 5 because the 'when' is terse and could be more specific about concrete trigger phrases.	4 / 5
Trigger Term Quality	The 'Trigger:' clause surfaces relevant keywords ('SDD research, external evidence, source-backed research') but these are framework-jargon-heavy and lack common synonyms or variations a user would naturally say. It is above 2 because multiple relevant terms are present, but not 4 due to missing natural phrasings.	3 / 5
Distinctiveness Conflict Risk	The 'SDD research' / 'source-backed research' niche is fairly distinct with clear triggers, though it risks minor overlap with sibling SDD phase skills (e.g. preproposal). It is not 5 because the SDD namespace contains related research/proposal skills that could overlap.	4 / 5
	Total	14 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_field	'metadata' should map string keys to string values	Warning

	Total	15 / 16 Passed

Repository: Gentleman-Programming/gentle-ai
Path: internal/assets/skills/sdd-research/SKILL.md
Commit: 71cf25f

Reviewed: 4 days 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.