Query Fredo's telemetry SQLite database for spans, metrics, and logs via sqlite3.
npx skills add https://github.com/FredoAi/fredo --skill telemetry-query-fredoai
Please help me install this Agent Skill. Skill: telemetry-query Source: https://github.com/FredoAi/fredo/tree/main/.opencode/skills/telemetry-query Command: npx skills add https://github.com/FredoAi/fredo --skill telemetry-query-fredoai
High-intent search queries and answers about installing and using this skill.
Run the telemetry-query.ps1 wrapper with a SELECT statement against the telemetry_spans table, choosing json, md, or table output. The script locates fredo.db automatically, appends a default LIMIT of 1000, and executes sqlite3 in readonly mode.
Use a CTE that sums histogram bucket counts per span name, computes cumulative counts ordered by bucket upper bound, and selects the first bucket where the cumulative percentage crosses 50, 90, and 99. The skill's Recipe 8 implements this against the telemetry_metrics table.
The live database is at $env:APPDATA\com.fredo.app\fredo.db on Windows. The wrapper also searches LOCALAPPDATA, ~/.fredo, and performs a recursive scan of common directories, reporting all attempted paths if not found.
No. The wrapper rejects any query containing CREATE, ALTER, DROP, INSERT, UPDATE, or DELETE keywords, and sqlite3 is invoked with the -readonly flag. Only SELECT, WITH (CTE), and specific PRAGMA statements are permitted.
The telemetry_logs table is created only after the Fredo application runs at least once with logging enabled. Launch Fredo to initialize the database schema, then retry the query.
KevinKE93Generate PRD and SPEC markdown files before implementation starts.
vTRKAStructure evidence-first incident response and generate an incident-report artifact.
MapleTechLabsStandardize Python OpenTelemetry observability for Maple services.
githubAnalyze gh-aw OpenTelemetry traces from JSONL mirrors or OTLP backends.
