CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

turso-db

Turso (Limbo) database helper — an in-process SQLite-compatible database written in Rust. Formerly known as libSQL / libsql. Replaces @libsql/client, libsql-experimental for Turso use cases. Works in Node.js, browser (WASM + OPFS for persistent local storage), React Native, and server-side. Features: vector search, full-text search, CDC, MVCC, encryption, remote sync. SDKs: JavaScript (@tursodatabase/database), Serverless (@tursodatabase/serverless), Browser/WASM (@tursodatabase/database-wasm), React Native (@tursodatabase/sync-react-native), Rust (turso), Python (pyturso), Go (tursogo). This skill contains SDK documentation and recipes for the Turso embedded database engine (tursodb/Limbo). Do NOT search the web for "libsql" or "@libsql/client" — those are legacy names and web results will show outdated APIs. Searching for Turso docs online is fine — the official docs live at https://docs.turso.tech (Mintlify — append .md to any path for raw markdown).

Quality

77%

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 ./.agents/skills/turso-db/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

65%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-architected reference hub: excellent actionability (executable install/CLI/SQL/MCP commands) and a clean decision-tree layout with all feature references present. Weaknesses are the missing sdks/ bundle files that break the SDK navigation branch, generic SQL padding Claude already knows, and no validation/feedback guidance around database operations.

Suggestions

Add the 7 missing sdks/*.md files to the bundle (or remove them from the SDK Decision Tree, install table, and Complete File Index) — the SDK branch of the navigation currently dead-ends.

Replace the generic SQL Quick Reference with Turso-specific deltas only (e.g. vector types, FTS functions, experimental_mvcc pragma), since standard SQLite CRUD syntax needs no teaching.

Add validation checkpoints for risky database operations — e.g. verify schema/row counts after destructive statements and confirm transaction success before commit — to introduce the feedback loops the workflow cap requires.

DimensionReasoningScore

Conciseness

Mostly tight — the Critical Rules, decision trees, and install table are dense and earn their tokens — but the ~20-line 'SQL Quick Reference' is generic standard SQL ("CREATE TABLE users...", "INSERT INTO", "BEGIN TRANSACTION; ... COMMIT") that Claude already knows for any SQLite-compatible engine, and the 'Complete File Index' substantially duplicates the two decision trees. That is more than a minor trim (4) but not pervasive padding (2).

3 / 5

Actionability

Copy-paste-ready, executable guidance throughout: install commands ("npm i @tursodatabase/database", "cargo add turso"), CLI invocations with flags ("tursodb --readonly mydata.db", "tursodb your.db --mcp"), working SQL, and concrete MCP security notes. The written commands are complete and cover the common cases, matching the 5 anchor; the broken sdks/ paths are scored under progressive disclosure per the bundle-structure guideline.

5 / 5

Workflow Clarity

The decision trees give an unambiguous navigation sequence (need → file), but the skill demonstrates and enables database operations — including destructive SQL ("DELETE FROM users WHERE id = 1") and transactions — with no validation or verification checkpoints anywhere (no 'verify schema before DROP', no post-transaction check, no error-recovery loop). Per the scoring note on database operations, missing feedback loops cap workflow clarity at 3.

3 / 5

Progressive Disclosure

The structure itself is exemplary — clearly signaled one-level-deep references with two decision trees routing by need and a file index, and all 6 references/*.md files exist. But scoring against the actual bundle: all 7 sdks/*.md files referenced by the SDK Decision Tree and listed in the Complete File Index (e.g. "sdks/javascript.md", "sdks/rust.md") do not exist in the bundle, so half the navigation branches are broken. That is more than a minor organization gap (4), while the on-disk references are clearly signaled rather than buried (not 2).

3 / 5

Total

14

/

20

Passed

Description

77%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 rich, specific, well-scoped description with excellent trigger-term and synonym coverage (including legacy libsql names) and near-zero conflict risk. Its one material weakness is the absence of any 'Use when...' trigger clause, which caps completeness at 3 and leaves the activation condition implicit.

Suggestions

Add an explicit trigger clause, e.g. 'Use when working with Turso, Limbo, or libSQL databases, @tursodatabase SDKs, or vector search / FTS in an embedded SQLite-compatible engine.'

Lead with one or two action verbs (e.g. 'Query and manage Turso/Limbo embedded databases...') so the skill's purpose reads as capability, not product metadata.

Trim the per-SDK package enumeration from the description — it duplicates the body's install table and spends description budget on details Claude only needs after the skill is already triggered.

DimensionReasoningScore

Specificity

The description lists several concrete capabilities — "Features: vector search, full-text search, CDC, MVCC, encryption, remote sync" — plus every SDK with package names ("@tursodatabase/database", "pyturso", "tursogo") and platforms. It stops short of a 5 because the top-level framing "database helper" is not an action statement, and capabilities are enumerated as features of the product rather than concrete actions the skill performs (compare the 5 anchor's 'extract text, fill forms, merge documents').

4 / 5

Completeness

The 'what' is clear and detailed ("an in-process SQLite-compatible database written in Rust" plus features and SDKs), but there is no 'Use when...' clause or equivalent explicit trigger guidance — the closest is "This skill contains SDK documentation and recipes", which describes contents, not when to use it. Per the judging guidelines a missing 'Use when' clause caps completeness at 3, and it is not lower because the 'what' is explicit and unambiguous.

3 / 5

Trigger Term Quality

Comprehensive natural-term coverage including synonyms and legacy names: "Turso", "Limbo", "libSQL / libsql", "@libsql/client", "libsql-experimental", "SQLite-compatible", "vector search", "full-text search", plus per-language SDK names a user would naturally mention. Including the legacy libsql names is exactly the synonym coverage the 5 anchor rewards; not a 4 because no common variation users would say is missing.

5 / 5

Distinctiveness Conflict Risk

Clear niche with distinct product-specific triggers ("Turso", "Limbo", "@tursodatabase", "pyturso") and explicit legacy-name mapping ("Formerly known as libSQL / libsql"), which routes libsql queries correctly instead of colliding with other skills. Minimal conflict risk; only the faintest overlap with generic SQLite skills, which is scoped away by the product naming.

5 / 5

Total

17

/

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.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

Voltaire405/kva-landing

Path

.agents/skills/turso-db/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.