CtrlK
BlogDocsLog inGet started
Tessl Logo

oracle

Claude-powered Q&A agent that answers questions about an indexed codebase using vector retrieval from ChromaDB. Supports /spec and /gaps commands.

57

Quality

65%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./workspace/skills/oracle/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

72%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 exceptionally concise and well-structured with appropriately split, real script references. Its weaknesses are instructional: it lacks concrete usage examples or a sequenced workflow that would let Claude execute the commands with confidence.

Suggestions

Add a brief usage example per command (e.g., sample '/spec src/auth.py' input→output) to lift actionability toward executable guidance.

Include a short sequenced workflow for handling a user query (retrieve → assemble context → answer with citations) with an explicit verification step that citations resolve to real paths/lines.

Optionally show the search_codebase tool call signature so the retrieval step is concrete rather than implied.

DimensionReasoningScore

Conciseness

The body is lean and efficient (~24 lines) with no padding or explanations of concepts Claude already knows; the one-line intro and capability/script lists all earn their place, matching the 'lean and efficient; every token earns its place' anchor.

5 / 5

Actionability

Provides some concrete guidance (command syntax '/spec [filepath]' and '/gaps' with described outputs, plus a script→class map) but no runnable code, example invocations, or specific execution steps; fits 'some concrete guidance but incomplete; missing key details' and below 4 which requires mostly-executable guidance.

3 / 5

Workflow Clarity

States one clear behavioral rule ('always retrieves before answering — never from memory alone') but lays out no sequenced multi-step workflow or validation checkpoints for the three command modes; the rule is cleaner than anchor 2's 'rough sequence, many gaps' yet lacks the explicit sequence of anchor 4.

3 / 5

Progressive Disclosure

Under 50 lines with well-organized sections (Capabilities, Scripts) and clearly-signaled, one-level-deep references to real bundle files (retriever.py, agent.py, spec.py, gaps.py, each mapped to its responsibility); per the simple-skill note this warrants 5 for clear, well-organized structure.

5 / 5

Total

16

/

20

Passed

Description

58%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 clearly conveys the skill's purpose and concrete commands but omits any 'Use when...' trigger guidance, capping completeness and limiting trigger-term quality. It is specific and fairly distinct, just missing the explicit activation context.

Suggestions

Add an explicit 'Use when...' clause naming natural user triggers (e.g., 'Use when the user asks questions about a codebase, wants a file spec, or needs to find coverage gaps').

Soften technical jargon toward natural phrasing users would actually say (e.g., 'ask questions about code', 'search the codebase') alongside ChromaDB.

Briefly note what /spec and /gaps produce so the actions read as fully concrete rather than terse command names.

DimensionReasoningScore

Specificity

Lists several concrete actions ('answers questions about an indexed codebase', 'vector retrieval from ChromaDB', '/spec', '/gaps'), matching the 'several specific actions; minor gaps' anchor; below 5 because the actions are named tersely without detailing what each command produces.

4 / 5

Completeness

Clearly states the 'what' (codebase Q&A via ChromaDB plus /spec and /gaps) but provides no 'when'/use-trigger guidance; per the rubric a missing 'Use when...' clause caps completeness at 3, so it cannot rise to 4.

3 / 5

Trigger Term Quality

Contains some relevant natural keywords ('Q&A', 'questions about... codebase') but leans on technical jargon ('vector retrieval from ChromaDB', 'indexed codebase') and omits common variations users would actually say; below 4 due to missing natural phrasing.

3 / 5

Distinctiveness Conflict Risk

Has a clear niche (indexed-codebase Q&A with named /spec and /gaps commands) yielding distinct triggers, but 'codebase Q&A' has minor overlap risk with broader code-search/explanation skills, fitting 'mostly distinct; minor overlap' rather than the minimal-conflict anchor 5.

4 / 5

Total

14

/

20

Passed

Validation

100%

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

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
tombrewsviews/repository-os-mvp
Reviewed

Table of Contents

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.