Menu
Analyze actual local Copilot skill activations without reading conversation content.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Use when a user wants to compare an arbitrary skill or path across models, quality, cost, latency, or routing.
Plan work too big for one agent session as a shared map of decision tickets on GitHub Issues or Jira, then resolve them one at a time until the route is clear. Use with /wayfinder when the user says "plan something too big for one session", "chart a map", "wayfind", "work the map", or "decision tickets". Always confirm tracker (GitHub vs Jira) and destination before writing. Adapted from mattpocock/skills wayfinder (MIT).
Create or resume a task handoff so a fresh agent can continue unfinished work. Persist via omp handoff CLI (never write handoff files yourself). Use when the user says /handoff, wants to pause mid-task, or resume prior work.
Single-owner execute-fix-verify loop for one clear task. Use with /ralph when one agent should keep going until evidence or blocker.
Adversarial QA pass that tests behavior, failures, and regressions. Use with /ultraqa after implementation when shallow checks are not enough.
Capture user-correction patterns from this session and, when a pattern recurs, draft a new project skill. Use at end of session, when the user signals wrap-up, or when invoked directly.
Based on SOC occupation classification
| name | history-analyze |
| description | Analyze actual local Copilot skill activations without reading conversation content. |
Use /history-analyze to request this local metadata-only history report.
Resolve the filters before execution. Accepted windows are 7d, 30d, 90d, and all, either
positionally or through --window/--since; accepted project scopes are current and all, either
positionally or through --project. Expand omitted filters to the defaults 30d and all.
Run exactly one normalized command, in this stable option order:
omp history analyze --window WINDOW --project SCOPE --json
Replace WINDOW and SCOPE with the resolved values. Never run the shorthand command after normalization.
Counts come only from actual skill tool execution-start events; never read conversation content,
prompt text, assistant text, or tool output, and never infer invocations from mentions. Preserve every warning
from the command. Usage attribution is session-level-only: never describe token, duration, AI-credit,
or model totals as per-skill values.
Usage totals cover skill sessions only. Present single-skill associations separately from the one shared-skill
bucket; never split or copy shared telemetry. Do not inspect session files directly.
Every numeric usage total includes metricSessions, the number of telemetry sessions that contributed
that specific metric; do not substitute the broader sessionsWithTelemetry count.
Require a successful schema-version-1 report before presentation. Present the history only after a successful schema-version-1 report. On schema or analyzer failure, stop without presenting history.
Do not print or paste raw JSON. Make the summary decision first: put the selected candidate from the
top-level skills array first, before supporting detail. The skills array ranks arbitrary observed skills from metadata only; it does not imply a supported benchmark, a semantic task, or failure cause.
Metadata-only history may rank observed skills or models, but must not claim semantic task or failure content. Then, within the available space:
sessionUsage value with its matching
metricSessions value;Before offering a handoff, use at most 12 compact bullets or lines and at most 2000 characters. Warnings may exceed the character bound only when necessary to preserve them verbatim; they still must not expose raw JSON. Omit lower-value detail rather than exceeding either bound. Copy numeric values exactly as returned. Do not round, abbreviate, rescale, or recalculate numeric values, and do not describe session-level usage as per-skill cost.
Resolve the next-step placeholders from the report and filters. If a candidate exists, Offer /skill-bench SELECTED_SKILL for the selected arbitrary skill. Always also Offer /skill-bench WINDOW SCOPE for guided discovery using the same filters.
Do not ask a consent question here and do not start a benchmark or live run from this skill. On silence, empty answer, non-answer, refusal, ambiguity, analyzer failure, or unavailable handoff, stop without starting a live run and print the valid handoff commands.