Issues

• Feature request: split cascade into estimator and flight-control partitions

  pulseengine/relay#575 · Filter by author cpetig opened on Oct 19, 2026

• Feature request: arena reclamation in falcon-cm-rt

  pulseengine/relay#574 · Filter by author cpetig opened on Oct 19, 2026

• scripts: read artifacts/*.yaml with rivet-yaml instead of serde_yaml; move falcon-quad.json to YAML

  pulseengine/relay#573 · Filter by author avrabe opened on Oct 16, 2026

• HITL wire binding carries no battery element, so a HITL host cannot arm the supervised component (#571)

  pulseengine/relay#572 · Filter by author avrabe opened on Oct 15, 2026
  1

• position and attitude leak their spilled parameter block every call — the 8 KiB arena exhausts after ~110 ticks

  bug
  pulseengine/relay#566 · Filter by author avrabe opened on Oct 14, 2026

• A latched failsafe is never cleared: the airframe cannot be re-armed after ANY failsafe, for the life of the process

  bug
  pulseengine/relay#564 · Filter by author avrabe opened on Oct 14, 2026
  3

• A sustained geofence breach actuates RTL for ONE 4 ms tick, then demotes it — the breach re-fires Failsafe every tick

  pulseengine/relay#562 · Filter by author avrabe opened on Oct 13, 2026

• Analyze if we can learn use tilize the dynamic drone failsafe

  pulseengine/relay#558 · Filter by author avrabe opened on Oct 11, 2026
  3

• OCI-P07's stage decomposition would LOSE the state-reset the monolith has — and the state it cannot clear is #270's root cause

  pulseengine/relay#553 · Filter by author avrabe opened on Oct 9, 2026
  1

• kani.yml runs all 47 legs on every main push: its scope check reads pull_request.base.sha, which is empty off-PR — four code-free merges cost falcon-v1.140.0 a day

  bug
  pulseengine/relay#548 · Filter by author avrabe opened on Oct 9, 2026
  3

• FV-FALCON-HOLD-001's verdict allows 2× HOLD-P01's bounds — the gate checks the final sample at the requirement, then the aggregates at twice it

  bug
  pulseengine/relay#546 · Filter by author avrabe opened on Oct 9, 2026
  2

• falcon-cm-rt: BumpArena::reset() at export entry can free caller-allocated arguments

  bug
  pulseengine/relay#544 · Filter by author cpetig opened on Oct 9, 2026
  3

• Two VERIFIED artifacts describe code that is not on main — and no step greps it, so the gate cannot see it

  pulseengine/relay#529 · Filter by author avrabe opened on Oct 7, 2026

• TRANSPORT-P01: the published wire binding accepts NaN and infinite motor commands — seam decision needed

  pulseengine/relay#527 · Filter by author avrabe opened on Oct 7, 2026
  1

• CI fleet: runs are starving (queued 30+ min)

  pulseengine/relay#514 · Filter by author github-actions opened on Oct 2, 2026
  4

• CI: 5 of 9 failed jobs today never ran — steps=0, no log blob, indistinguishable from a real failure

  pulseengine/relay#508 · Filter by author avrabe opened on Oct 1, 2026
  1

• TRANSPORT-P01 quotes a capture that is not in the repo — and its stated header size contradicts its own field list

  pulseengine/relay#503 · Filter by author avrabe opened on Oct 1, 2026

• v1.141 clean-room review: all five artifacts refuted — the measurements reproduce, the evidence statements do not

  pulseengine/relay#499 · Filter by author avrabe opened on Oct 1, 2026

• v1.140 clean-room review: five blocking findings — ENDURANCE's 1 h gz hold never ran, FLEET-002 cannot fail

  pulseengine/relay#497 · Filter by author avrabe opened on Oct 1, 2026

• The flight seam is the ONE WIT we hand-write — spar cannot express it, and nothing gates its drift

  pulseengine/relay#492 · Filter by author avrabe opened on Sep 30, 2026

• MockPhysics computes specific force BEFORE the ground clamp — the same diff fixed exactly this in SimBackend

  pulseengine/relay#485 · Filter by author avrabe opened on Sep 23, 2026

• FV-FALCON-HOLD-001 cites a test that cannot fail for the branch it names (mutation-tested)

  pulseengine/relay#483 · Filter by author avrabe opened on Sep 23, 2026
  2

• flying() gates the estimator's attitude reference on hover_thrust — and the shipped component hardcodes 0.5

  pulseengine/relay#481 · Filter by author avrabe opened on Sep 23, 2026

• The 'hover' scenario diverges (74.6 m at 25 s, 202 m at 60 s) — a legacy path superseded by 'flightcore', still shipped as a scenario

  pulseengine/relay#477 · Filter by author avrabe opened on Sep 23, 2026

• The starvation check keys on RUN status — a matrix run whose jobs are all being picked up alarms as 'queued 30+ min' (#447 was a false positive)

  pulseengine/relay#449 · Filter by author avrabe opened on Sep 17, 2026
  1