Governed Agent Harness — process metrics

Process telemetry of governed runs, read through the harness application layer. Values are aggregated per run and per project only; the harness produces no per-person indicators. Every metric carries its data quality (OBSERVED, REPORTED, DERIVED, ESTIMATED, NOT_AVAILABLE); token and cost values stay NOT_AVAILABLE unless the provider reports them.

27 run(s) across 17 project(s): agent-results, bdd, broken-baseline, clarification, corrections, delivery, friction, gitlab, integrity, ladder, later-change, memory, node, quickstart, review-exception, tdd, traceability.

Gate states

Latest gate evaluation per run
FAILEDFAILED: 2 runs2 runsNO_GATENO_GATE: 4 runs4 runsPASSEDPASSED: 21 runs21 runs
Latest gate evaluation per run
Gate stateRuns
FAILED2
NO_GATE4
PASSED21
Execution state per run
Run stateRuns
BLOCKED4
CANCELLED1
FAILED5
PASSED17

Duration per phase

Median wall-clock duration per phase attempt
INTENTINTENT: 26 ms26 msDISCOVERYDISCOVERY: 597 ms597 msSPECIFICATIONSPECIFICATION: 32 ms32 msPLANNINGPLANNING: 17 ms17 msIMPLEMENTATIONIMPLEMENTATION: 104 ms104 msVERIFICATIONVERIFICATION: 1,971 ms1,971 msINDEPENDENT_REVIEWINDEPENDENT_REVIEW: 13.5 ms13.5 msDECISIONDECISION: 20 ms20 msCLOSURECLOSURE: 102 ms102 ms
Median duration per phase attempt, all runs
PhaseMedian ms
INTENT26
DISCOVERY597
SPECIFICATION32
PLANNING17
IMPLEMENTATION104
VERIFICATION1,971
INDEPENDENT_REVIEW13.5
DECISION20
CLOSURE102

Attempts, corrections and human wait per run

Per-run process metrics (no actor dimension)
ProjectRunStateGateImpl. attemptsCorrection cyclesReview cyclesNon-passed validationsChangeSetsHuman decisionsHuman wait msEventsChain valid
agent-results4a978526c551FAILEDPASSED1012114,455130yes
bddf46bc91a13d6BLOCKEDPASSED101110—151yes
broken-baseline1ec9bb15e734PASSEDPASSED1014111,72688yes
clarification75a499abd25cCANCELLED—000000—27yes
clarification3367a3fe35c0PASSEDPASSED1011112,027156yes
clarification5951219a6fd1PASSEDPASSED1011111,987167yes
corrections9566f9401a6eBLOCKEDPASSED101110—64yes
correctionsda4d9ba55fe1FAILED—320910—117yes
corrections4679069cb57aPASSEDPASSED2114212,574106yes
delivery4d0c139690cbPASSEDPASSED1011112,080140yes
friction06bad6c3257ePASSEDPASSED1013111,951135yes
friction838f37508f48PASSEDPASSED1013114145yes
gitlab85e347d65504BLOCKEDPASSED101110—55yes
integrityf79fb345784aPASSEDPASSED1011116,07969no
ladderda9eb5363ca7BLOCKEDPASSED101210—136yes
ladder76e1f92c6a1bFAILED—100210—103yes
ladder7805448011cfPASSEDPASSED1011114,039168yes
later-change342e14218533PASSEDPASSED2123321,823116yes
memoryd1bb8cbfcd03PASSEDPASSED1011112,53173yes
memory9802442f9f4fPASSEDPASSED1011112,56570yes
memoryfb99a29464f3PASSEDPASSED1011112,53069yes
node5be43aa3a748PASSEDPASSED1012112,024128yes
quickstart6414c678a191PASSEDPASSED1010112,065135yes
review-exceptiond9f1d96ed3c9PASSEDFAILED1011112,73872yes
tdd368a37e1733cFAILED—100210—104yes
tddf10e6722daedPASSEDPASSED1011113,011142yes
traceability22f905aec0a1FAILEDFAILED1011115,027126yes

Authorized exceptions

1 run(s) closed with APPROVE_EXCEPTION
ProjectRunGate at decisionRecorded rationale
review-exceptiond9f1d96ed3c9FAILEDDemonstration fixture: the value is not a real credential
Latest human decision per run
Final decisionRuns
APPROVE16
APPROVE_EXCEPTION1
NONE8
REJECT2

Data quality per metric

Quality label reported by the harness for each metric
MetricUnitQuality (runs)DefinitionSource
agent.invocationscountOBSERVED×27Number of persisted agent invocation records.agent invocation records
agent.transient_retriescountOBSERVED×27Command-provider calls repeated after a transient failure (runtime.providerRetries).event store
agent.unsupported_claimscountDERIVED×27Findings agent.unsupported-claim: the agent reported success and verification failed.finding records
changeset.filescountDERIVED×27Unique paths appearing in persisted ChangeSets.ChangeSet records
changesets.countcountOBSERVED×27Number of distinct persisted ChangeSet records.ChangeSet records
correction.cyclescountOBSERVED×27Count of authorized transitions back to IMPLEMENTATION: REQUEST_CHANGES decisions and automatic corrections after a failed VERIFICATION.event store
correction.verification_cyclescountOBSERVED×27Automatic corrections after a failed VERIFICATION (runtime.verificationCorrections).event store
cost.usdUSDNOT_AVAILABLE×26, REPORTED×1Sum of provider-reported costs; never inferred from tokens.resource usage records
duration.agent_msmsDERIVED×27Sum of persisted agent invocation durations.agent invocation records
duration.human_wait_msmsDERIVED×19, NOT_AVAILABLE×8Elapsed time from the latest gate evaluation to the latest human decision.gate and decision events
duration.phase_msmsDERIVED×27Sum of completed phase wall-clock durations; parallel phases would be double-counted.phase records
duration.tool_msmsDERIVED×27Sum of persisted tool invocation durations.tool invocation records
duration.total_msmsOBSERVED×27Wall-clock time between the first and last persisted execution event.event store
human.decisionscountOBSERVED×27Number of persisted human decisions.decision records
human.interactionscountOBSERVED×27Events a person caused on the run: decisions, clarification answers, plan, acceptance and preflight decisions, budget raises, quarantines, contract confirmations and attached evidence.run events with a human actor
implementation.attemptscountDERIVED×27Count of IMPLEMENTATION phase-start events.event store
replanning.countcountOBSERVED×27Count of plan replacement events after the first accepted plan.event store
review.cyclescountDERIVED×27Count of INDEPENDENT_REVIEW phase-start events.event store
tokens.inputtokensNOT_AVAILABLE×24, REPORTED×3Sum of provider-reported input_tokens; no estimation is performed.resource usage records
tokens.outputtokensNOT_AVAILABLE×24, REPORTED×3Sum of provider-reported output_tokens; no estimation is performed.resource usage records
tokens.reasoningtokensNOT_AVAILABLE×27Sum of provider-reported reasoning_tokens; no estimation is performed.resource usage records
tool.invocationscountOBSERVED×27Number of persisted tool invocation records.tool invocation records
validation.non_passedcountDERIVED×27Validation results whose normalized status is not PASSED.validation records