CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

reading-source-code

Workflow for reading repository code with `exploreRepo` and `discoverRepos`, covering when to ask which repositories to search and how to report findings and errors. Load it before answering from repository code.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Critical

Do not install without reviewing

SKILL.md
Quality
Evals
Security

Quality

Content

92%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.

An exemplary lean instruction-only skill: three clearly signaled behavioral directives, each with concrete conditions, example phrasings, and error-recovery guidance, and no wasted tokens. The only room for improvement is making a few implicit specifics (e.g., example error messages or candidate-offering phrasing) explicit.

DimensionReasoningScore

Conciseness

Every sentence carries tool-specific guidance Claude could not infer ("Each repository is searched separately, so covering the whole set is slow and rarely needed"; "try `grep` before saying nothing matched") with zero padding or explanation of known concepts, matching the lean-and-efficient anchor.

5 / 5

Actionability

The guidance is concrete and executable at the behavioral level with example utterances ("tell them how many repositories `discoverRepos` lists and ask which one or few to focus on"; "Three files reference `OrderStatus`: …"; "fix it and retry: the error says how"); minor gaps (example error shapes, phrasing for offering candidates) keep it just below fully copy-paste-ready.

4 / 5

Workflow Clarity

A simple under-50-line skill whose behavior is unambiguous, with an explicit confirmation checkpoint before the batch operation ("Ask before scanning many repositories... Search them all once the user confirms") and an explicit error-recovery loop ("When a command fails... fix it and retry"), satisfying the simple-skill exception with validation present.

5 / 5

Progressive Disclosure

The skill is under 50 lines, needs no external references, and is well organized into a header plus three clearly signaled bolded directive sections, which merits a 5 under the under-50-lines guideline; no bundle files exist to verify.

5 / 5

Total

19

/

20

Passed

Description

67%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 solid description that states concrete tool-specific capabilities and includes an explicit load-trigger clause. Its main weakness is trigger-term coverage: it uses tool names and 'repository code' but omits the natural phrases (codebase, search the code, find where something is implemented) a user would most likely say.

Suggestions

Add natural trigger phrases to the description, e.g. "Use when the user asks to find where something is implemented, search the codebase, or read repository code".

Surface the term "source code" and "codebase" in the description itself (not just the skill name) so common user phrasings match.

Briefly state the outcome delivered (locating code that answers a question) to sharpen the "what" beyond naming the tools.

DimensionReasoningScore

Specificity

"reading repository code with `exploreRepo` and `discoverRepos`", "when to ask which repositories to search", and "how to report findings and errors" name several distinct concrete actions, matching the anchor for several specific actions with minor gaps; it falls short of 5 because coverage is not comprehensive (e.g., the search/locator behavior is not described).

4 / 5

Completeness

It answers both parts: the "what" ("Workflow for reading repository code with `exploreRepo` and `discoverRepos`...") and an explicit "when" ("Load it before answering from repository code"); the when is explicit but lacks concrete user-side trigger phrases, so it fits anchor 4 rather than 5.

4 / 5

Trigger Term Quality

"repository code" and "repositories to search" are relevant keywords, but common natural phrasings users would actually say ("codebase", "search the code", "find where X is implemented", "source code") are missing, matching the anchor for some relevant keywords missing common variations.

3 / 5

Distinctiveness Conflict Risk

Naming the specific tools "`exploreRepo` and `discoverRepos`" carves a clear niche with minor overlap risk against generic code-search skills; it is not a 5 because no concrete trigger phrases fully disambiguate it from adjacent repository-search skills.

4 / 5

Total

15

/

20

Passed

Validation

93%

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

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

lightdash/lightdash

Path

packages/backend/src/ee/services/ai/skills/builtInSkills/reading-source-code/SKILL.md

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.