CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

grounding

Establish what is actually true and current for the surface you are about to change — not just search. Grounding = locate the authoritative source and reconcile sources that disagree (code vs doc, migration vs schema, memory vs current code, live vs archived), not take the first hit. Use before any grep/find/explore of the codebase or docs, when deciding which of several definitions is the real one, or when a doc or memory conflicts with the code. Covers the source-of-truth hierarchy, reconciliation discipline, scoped ripgrep, and the docs/tags.yml + docsearch.py index.

Quality

85%

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

SKILL.md
Quality
Evals
Security

Quality

Content

86%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 skill body: Grida-specific facts only, executable commands, and a well-structured decision framework with the docsearch script correctly kept one level deep. The only refinements are trimming the meta seed note and stale-prone file counts, and adding slightly more detail on the conflict-escalation step.

DimensionReasoningScore

Conciseness

The body is dense and assumes competence (no explaining what ripgrep or frontmatter is), with nearly every line carrying a Grida-specific fact; minor trimmable bits remain — the "Intentional seed" meta blockquote and the stale-prone "~360 markdown files" count — though the "retired May 2026" date is properly confined to the deprecation-style "Dead-tree traps" section.

4 / 5

Actionability

Fully executable, copy-paste-ready commands throughout ("rg PATTERN -g '!**/docs/_history/**'", "uv run $S find --tag figma --tag wg", "uv run $S show wg/feat-fig/glossary/fig.kiwi.md") plus concrete tie-breaker criteria ("git log -1 recency, what the running entrypoint imports, what tests assert") and doc-trust checks ("draft: true", "doc_tasks:") covering the common cases.

5 / 5

Workflow Clarity

The decision framework is well sequenced — name the surface, pick authority from the per-surface hierarchy, reconcile disagreement with explicit tie-breakers, exclude dead trees, check doc freshness signals — but the final "surface a material conflict to the user" step has no recovery detail, leaving a minor checkpoint gap relative to the level-5 anchor.

4 / 5

Progressive Disclosure

Verified against the actual bundle: the single referenced script (scripts/docsearch.py) exists, is one level deep, and its full usage docs live in the script's own docstring while the body inlines only the essential invocations; sections are clearly organized and the ai-models cross-skill link is well signaled.

5 / 5

Total

18

/

20

Passed

Description

80%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 with a clear what/when structure and concrete trigger conditions. Its main weakness is trigger breadth: "before any grep/find/explore" claims territory far beyond the source-of-truth reconciliation niche, raising conflict risk with generic search skills.

Suggestions

Narrow the "Use before any grep/find/explore of the codebase or docs" trigger to situations involving conflicting or possibly-stale sources (e.g. "Use when sources disagree or may be stale: doc vs code, migration vs schema, memory vs current code") so it does not fire on ordinary code exploration.

Add the natural synonyms users would actually say — "source of truth", "search", "verify" — to improve trigger term coverage.

Replace the abstract phrase "reconciliation discipline" with the concrete action it names (e.g. "decide which source wins and why").

DimensionReasoningScore

Specificity

Lists several concrete actions — "locate the authoritative source and reconcile sources that disagree", "scoped ripgrep", "the docs/tags.yml + docsearch.py index" — but mixes in abstract framing like "reconciliation discipline", keeping it below the comprehensive pure-action level 5.

4 / 5

Completeness

Clearly answers both: the "what" ("Establish what is actually true and current… locate the authoritative source and reconcile sources that disagree") and an explicit "Use when…" clause with three concrete trigger conditions, matching the top anchor.

5 / 5

Trigger Term Quality

Natural trigger phrases are present ("Use before any grep/find/explore of the codebase or docs", "when a doc or memory conflicts with the code", "which of several definitions is the real one"), but common synonyms such as "search", "source of truth", or "verify" are missing.

4 / 5

Distinctiveness Conflict Risk

The grounding/source-of-truth niche itself is distinct, but the trigger "Use before any grep/find/explore of the codebase or docs" is broad enough to fire on nearly any code-exploration task, creating real overlap with general search/research skills; it is not level 4 because the overlap risk is more than minor.

3 / 5

Total

16

/

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

relative_links

Relative link issues: 1 suspicious

Warning

Total

15

/

16

Passed

Repository

gridaco/grida

Path

.agents/skills/grounding/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.