Run History
Read the workspace's runs through the native run.list service and the
installed catalog through runx.skill.inspect. The method computes the report
itself; it does not ask an agent to invent commands, launch subprocesses, or
transcribe counts.
Use it for operational questions about recent Runx activity, failed or blocked
runs, paused runs, frequently run skills, and skill entries without fixtures or
inline test cases. Use receipt-audit or diagnose-skill-run to
inspect one suspicious run, and least-privilege when receipt-backed authority
evidence is available for grant review; actual usage and policy semantics are still needed before attenuation.
Evidence boundary
- Live execution reads only the native run list and catalog projections. These
services read the same run records and catalog parser as the Runx CLI
without routing back through the CLI as a subprocess.
- Run reads are bounded to 1,000 matching runs. Reaching that bound is explicit;
the reported rates never claim whole-ledger coverage. An explicit
as_of is
also passed as the native upper bound, excluding later runs.
runs and catalog_items are replay inputs for tests and controlled
analysis, and must be supplied together; replay never silently mixes caller
data with live native state. Supply already-windowed projections and omit
period/as_of for replay; time filtering belongs to run.list, not a second
JavaScript history engine. Skill scope is applied before the replay limit,
preserving the original supplied order.
- Empty history or a native list containing unreadable records returns
needs_more_evidence. Unreadable run ids, update times and reasons remain in
runs.unreadable; they count toward the native limit and are excluded from
outcome rates. Native skill/time filters cannot attribute unreadable records
to a narrower selection, so filtered reports do not establish their absence.
completed_rate is the share of sealed runs whose outcome is completed.
refusal_rate counts sealed runs that were blocked or declined. It is
reported as an observation, not automatically treated as a defect.
- Catalog coverage is the number of native catalog entries declaring at least
one fixture or inline test case. It is not a maturity grade.
Output
history_report contains:
- the resolved query and source labels (
run.list, runx.skill.inspect, or
supplied_replay);
- sealed and paused run counts, outcome counts, completed/refusal rates;
- the most frequently run skills;
- catalog entry and test-coverage counts;
- bounded recommendations routed to
diagnose-skill-run, receipt-audit, or
skill-build test;
- limitations when native projections cannot support a stronger claim.
This skill never executes another skill, changes a grant, or changes a run.
1---2name: run-history3description: Read the workspace's runs and skill catalog, then return deterministic run outcomes, paused runs, catalog test coverage, and governance follow-ups without model-authored metrics.4metadata:5 source: https://skillmd.com/skills/runxhq/run-history6---78# Run History910Read the workspace's runs through the native `run.list` service and the11installed catalog through `runx.skill.inspect`. The method computes the report12itself; it does not ask an agent to invent commands, launch subprocesses, or13transcribe counts.1415Use it for operational questions about recent Runx activity, failed or blocked16runs, paused runs, frequently run skills, and skill entries without fixtures or17inline test cases. Use `receipt-audit` or `diagnose-skill-run` to18inspect one suspicious run, and `least-privilege` when receipt-backed authority19evidence is available for grant review; actual usage and policy semantics are still needed before attenuation.2021## Evidence boundary2223- Live execution reads only the native run list and catalog projections. These24 services read the same run records and catalog parser as the Runx CLI25 without routing back through the CLI as a subprocess.26- Run reads are bounded to 1,000 matching runs. Reaching that bound is explicit;27 the reported rates never claim whole-ledger coverage. An explicit `as_of` is28 also passed as the native upper bound, excluding later runs.29- `runs` and `catalog_items` are replay inputs for tests and controlled30 analysis, and must be supplied together; replay never silently mixes caller31 data with live native state. Supply already-windowed projections and omit32 `period`/`as_of` for replay; time filtering belongs to `run.list`, not a second33 JavaScript history engine. Skill scope is applied before the replay limit,34 preserving the original supplied order.35- Empty history or a native list containing unreadable records returns36 `needs_more_evidence`. Unreadable run ids, update times and reasons remain in37 `runs.unreadable`; they count toward the native limit and are excluded from38 outcome rates. Native skill/time filters cannot attribute unreadable records39 to a narrower selection, so filtered reports do not establish their absence.40- `completed_rate` is the share of sealed runs whose outcome is `completed`.41- `refusal_rate` counts sealed runs that were `blocked` or `declined`. It is42 reported as an observation, not automatically treated as a defect.43- Catalog coverage is the number of native catalog entries declaring at least44 one fixture or inline test case. It is not a maturity grade.4546## Output4748`history_report` contains:4950- the resolved query and source labels (`run.list`, `runx.skill.inspect`, or51 `supplied_replay`);52- sealed and paused run counts, outcome counts, completed/refusal rates;53- the most frequently run skills;54- catalog entry and test-coverage counts;55- bounded recommendations routed to `diagnose-skill-run`, `receipt-audit`, or56 `skill-build test`;57- limitations when native projections cannot support a stronger claim.5859This skill never executes another skill, changes a grant, or changes a run.