Robot-code migration onto ExcaLib v2 (stacked on #47) by YehudaRothstein · Pull Request #48 · ExcaliburFRC/Rebuilt2026

YehudaRothstein · 2026-07-02T15:57:19Z

What this is

Stacked on #47. Every subsystem, the superstructure, and the drivetrain rebuilt on the ExcaLib v2 archetypes — the proof that the template layer works. This is a refactor onto a better foundation, not a behavior change: all tuning data (interpolation maps, setpoints, tolerances, current limits) is ported verbatim with the v1→v2 unit derivations documented inline. The old library remains in-tree and compiling (removal only after explicit approval).

Migration map

Commit	Subsystem	Before	After
B1	Transport	2× software-PID FlyWheel, polled enum field	2× `VelocityMechanism` (onboard 1 kHz loops) + IDLE/TRANSPORT machine; shooter-ready gating preserved
B2	Intake	hand-rolled Arm control law (error/dt velocity hack), pump via timeouts	`PositionalMechanism` (MotionMagic @ v1's 1.5 rad/s + 1.5 V clamp, arm-cosine kG, device soft limits) + `RollerMechanism`; OPEN/CLOSE/PUMP machine (PUMP oscillation preserved)
B3	Shooter	hood hand-PID w/ magic kS, turret software profile, flywheel software PID	hood/turret `PositionalMechanism` (turret continuous-wrap in −5.4..1.5 rad; hood trench-aware max), flywheel `VelocityMechanism`; maps, 3-iteration shot lead, radial-velocity comp, BallCounter, LEDs, shooterReady — verbatim
B4	Superstructure	ParallelCommandGroup-of-InstantCommands mutating fields	`Superstructure<RobotState>` goal machine, one fan-out; trigger wiring ported 1:1 (zone stubs kept stubbed)
B5	Swerve + vision	hand-rolled modules (software PID), MegaTag1 in periodic	CTRE `SwerveDrivetrain` via hand-authored `SwerveConfig`; MegaTag2 with turret-corrected orientation + turret→robot pose transform + isFast gate

🟥 Behavior deltas — read before driving

All migrated mechanisms run VOLTAGE control mode, not TorqueCurrentFOC: the ported gains are v1's volts-based gains (math-equivalent, now at 1 kHz onboard). The FOC switch is a one-line config change per mechanism after a SysId session — gains do not transfer between modes.
Swerve needs on-robot verification before competition: module inversions, CANcoder offsets (assumed burned into devices, as v1 read raw absolute angles), coupling ratio (set 0), gain re-tune for the CTRE stack. Bench-test checklist in the SwerveConfig Javadoc.
Named auto commands are registered again (registerCommands() was commented out on main — audit P-04): autos will now actually intake/shoot. Verify auto behavior is desired before fielding.
Hood kS is now symmetric 0.3125 V (v1 hand-coded +0.375/−0.25) — ±0.06 V difference.
Signals refresh before the scheduler (v1: after) — removes ~20 ms sensor latency (strictly better, but measurable in tuning).
Deadband is now rescaled (no output step at the deadband edge).
Turret-aligned trigger compares to the commanded goal (v1 recomputed the target in the trigger).

Testing evidence

✅ ./gradlew build green at every commit.
✅ simulateJava boots clean: robot program starts, DogLog + SignalLogger live, no exceptions, no loop overruns over a 60 s soak (the only console error is v1's pre-existing "DS alliance empty" report, present on main too).
✅ All mechanisms have sim physics models — hood/turret/four-bar/flywheels/drivetrain move in the sim GUI.
⏳ Not yet done (needs the robot): SysId runs, FOC gain derivation, swerve constant verification, on-field shooter validation.

Follow-ups

🟥 On-robot swerve verification session + SysId/FOC tuning session (see deltas 1–2).
🟧 Fuse the intake/hood/turret CANcoders as device feedback (needs offset calibration; today they seed position at boot, v1-style).
🟧 Delete ExcaLib v1 + Monologue + leftover dead robot code — separate PR after approval (rule 7).
🟨 Re-enable the real field-zone triggers (stubbed on main; kept stubbed here on purpose).
🟨 Raise MAX_SPEED_MPS/MAX_OMEGA from the v1 test values (2 m/s / 1.5 rad/s) once the drivetrain is verified.

🤖 Generated with Claude Code

Update: API style pass + full v1 feature parity

Two follow-up requests applied on this branch:

No static factory methods (team style): everything is constructed with new and configured with instance methods — new Gains(kP,kI,kD,kS,kV,kA), new MotionConstraints(...) (3-arg = trapezoidal, 2-arg = Expo), new CurrentBudget(stator, supply), new MechanismConfig(name, id), new StateMachine<>(name, initial) (no builder — declarations chain on the instance, initial state fires on first tick), and drive commands live on SwerveSubsystem (swerve.fieldCentricDriveCommand(...)) instead of a DriveCommands helper class.
Full v1 feature parity — see docs/V1_TO_V2_MAPPING.md for the complete v1-class → v2 table. Highlights: SwerveSubsystem now has every v1 Swerve capability (heading-locked drive, pathfind-to-pose / then-follow-path / with-driver-override, reset heading/pose, coast, sawTagRecently, Field2d, drivetrain SysId); new LedStrip (WPILib LEDPattern), LoopTimer, alliance pose types on AllianceFlip, CAN-bus health in FaultReporter. Robot code no longer uses any v1 utilities (only Monologue + PeriodicScheduler remain until v1 deletion).

Re-verified after the changes: build green, 9/9 unit tests, sim boots with no exceptions.

Update: ExcaLib v1 removed ✅ (approved)

With the migration proven, the old library was deleted:

frc.excalib package removed (53 files: motor wrappers, mechanisms, swerve, slam/aurora, imu, math, utilities).
Constants.java: dead SwerveConstants inner class dropped (replaced by SwerveConfig); v1 + now-unused imports removed.
Robot.java: PeriodicScheduler.run() and TalonFXMotor.refreshAll() dropped — nothing registered on either (all motors are excalib2 Motor via SignalHub).
Vestigial .gitmodules and stale MOTOR_LIBRARY_CHANGES.md removed.
Docs updated (CLAUDE.md, README, manual, v1→v2 mapping).

Verified: ./gradlew build green (compile + unit tests); simulateJava boots clean with no exceptions; zero frc.excalib references remain in src. Only follow-up: retiring Monologue in favor of DogLog (separate vendordep, tracked independently).

…state machine Drum + belt become VelocityMechanisms (onboard 1 kHz velocity loops, verified configs, DogLog telemetry, sim models). Two-state machine (IDLE/TRANSPORT) replaces the polled enum field; shooter-ready gating preserved. v1 gains/setpoints ported in legacy units (documented conversion) for behavior parity; VOLTAGE mode until the FOC SysId session. Robot loop now refreshes SignalHub before the scheduler and polls FaultReporter; Telemetry.init on boot. Public API (setStateCommand/atPositionTrigger/manualTransport) kept — Superstructure untouched in this commit. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Four-bar becomes a PositionalMechanism (leader/follower Krakens, MotionMagic bounded at v1's 1.5 rad/s, arm-cosine kG, v1 1.5 V output clamp preserved via peak-voltage config, device-side soft limits) and the roller a RollerMechanism. OPEN/CLOSE/PUMP state machine replaces the polled enum; PUMP keeps the v1 oscillation (1.6 rad <-> 0 rad at 0.5 s, roller at 0.8 duty). Position still seeded from the v1 stow constant; fusing the existing CANcoder is a flagged follow-up (needs on-robot offset calibration). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

…hine Turret and hood become PositionalMechanisms (turret: continuous-wrap goals inside -5.4..1.5 rad with v1's trapezoid profile; hood: plain position loop with trench-aware dynamic max), flywheel a two-Kraken VelocityMechanism. All v1 competition logic preserved verbatim: interpolation maps, 3-iteration shot lead, radial-velocity compensation, EMA-filtered readiness, BallCounter, LED blink, TurretOffsetGetter feed for vision. Ratios derived from v1 conversion factors (documented); gains ported V/rad -> V/rot in VOLTAGE mode (FOC after SysId). Deltas called out: symmetric hood kS 0.3125 (was +0.375/-0.25), turret-aligned now compares to the commanded goal. Superstructure adapted to the new coast/manual-shooting API. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

…ister auto commands Superstructure now extends excalib2's Superstructure<RobotState>: each goal fans out to the intake/shooter/transport state machines in one onEnter, replacing the ParallelCommandGroup-of-InstantCommands. Trigger wiring preserved 1:1 (including the stubbed zone triggers, kept stubbed on purpose); dead ledStateFor and 15 copy-paste NT getters replaced by DogLog lines. Also re-enables RobotContainer.registerCommands() (audit P-04: autos referenced idle/shoot/intake named commands that were never registered and silently no-op'd). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

…RE swerve + MT2) Hand-authored Tuner-style constants in SwerveConfig, derived from v1 values (drive 5.27:1, steer 26.09:1, 2 in wheels, FusedCANcoder with device-burned offsets, v1 gains converted to per-rotation volts). RobotContainer drives via DriveCommands.fieldCentric (deadband now rescaled); Superstructure decoupled from the drivetrain type (takes pose/speeds suppliers). Vision moves from MT1-in-periodic to the MegaTag2 path: turret-corrected orientation feed, turret-frame -> robot-frame pose transform (v1 turretToRobot math), TurretOffsetGetter isFast rejection gate, distance/tag-count std-dev scaling. PathPlanner AutoBuilder wired with wheel-force feedforwards. FaultReporter skips fault scans in sim (signals not modeled -> console spam). ⚠ Needs on-robot verification before competition: module inversions, CANcoder offsets, coupling ratio (set 0), drive/steer gain re-tune. Old ExcaLib v1 remains in-tree and compiling (removal only after approval). Verified: build green; simulateJava boots clean (no exceptions, no loop overruns). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

coderabbitai · 2026-07-02T15:57:32Z

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: f3e9f3a5-6ad5-470e-86bd-06edc0b71bcf

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch feature/robot-migration

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands.}

Copilot

Pull request overview

Migrates the robot implementation (frc.robot) onto ExcaLib v2 (frc.excalib2) archetypes (mechanisms + state machines), including the superstructure goal machine and CTRE swerve/vision integration, while keeping ExcaLib v1 in-tree and compiling for reference.

Changes:

Rebuilt Intake / Shooter / Transport as ExcaLib v2 PositionalMechanism / VelocityMechanism / RollerMechanism + StateMachine-driven subsystems, with tuning data and key targeting logic preserved.
Introduced a hand-authored CTRE swerve SwerveConfig and added turret-mounted Limelight MegaTag2 pose transformation support.
Updated robot lifecycle to initialize v2 telemetry and refresh/optimize Phoenix status signals ahead of command execution; re-enabled PathPlanner named-command registration.

Reviewed changes

Copilot reviewed 17 out of 17 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
src/main/java/frc/robot/SwerveConfig.java	New CTRE swerve constants + Limelight MT2 wiring + turret-pose→robot-pose transform hook.
src/main/java/frc/robot/superstructure/Superstructure.java	Refactors superstructure into v2 goal machine that fans out subsystem state requests.
src/main/java/frc/robot/subsystems/transport/TransportConstants.java	Ports transport tuning into v2 gains/budgets + adds sim models and legacy-unit derivations.
src/main/java/frc/robot/subsystems/transport/Transport.java	Rebuilds transport as two v2 `VelocityMechanism`s controlled by a 2-state machine with shooter-ready gating.
src/main/java/frc/robot/subsystems/shooter/ShooterConstants.java	Ports shooter ratios/gains/constraints into v2 units, adds budgets + sim models.
src/main/java/frc/robot/subsystems/shooter/Shooter.java	Rebuilds shooter as v2 mechanisms + state machine; preserves interpolation maps and lead/compensation math; adds DogLog telemetry.
src/main/java/frc/robot/subsystems/intake/IntakeConstants.java	Ports intake geometry/gains/motion/budgets + sim models; documents v1→v2 derivations.
src/main/java/frc/robot/subsystems/intake/Intake.java	Rebuilds intake as v2 four-bar + roller mechanisms with OPEN/CLOSE/PUMP state machine.
src/main/java/frc/robot/RobotContainer.java	Switches to v2 swerve, restores named-command registration, updates teleop drive command wiring.
src/main/java/frc/robot/Robot.java	Initializes v2 telemetry, optimizes signals at boot, refreshes signals + polls faults before commands.
src/main/java/frc/excalib2/telemetry/FaultReporter.java	Skips fault polling in simulation to avoid stale-CAN spam.
src/main/java/frc/excalib2/swerve/SwerveSubsystem.java	Adds camera-pose→robot-pose transform support for non-chassis-mounted cameras.
src/main/java/frc/excalib2/mechanisms/VelocityMechanism.java	Selects between MotionMagicVelocity vs plain velocity closed-loop based on config motion constraints.
src/main/java/frc/excalib2/mechanisms/PositionalMechanism.java	Adds plain position closed-loop path when no motion constraints are configured.
src/main/java/frc/excalib2/device/ExcaTalonFX.java	Adds plain position/velocity control requests alongside MotionMagic-based requests.
README.md	Updates repo layout documentation to reflect ExcaLib v2 structure and migration state.
CLAUDE.md	Updates contributor/agent guide to reflect v2 architecture, control conventions, loop order, and tests.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+                    .rotorToMechanismRatio(FLYWHEEL_ROTOR_PER_MECHANISM)
+                    .gains(FLYWHEEL_GAINS, FLYWHEEL_SIM_GAINS)
+                    .currentBudget(FLYWHEEL_BUDGET)
+                    .tolerance(Radians.of(FLYWHEEL_TOLERANCE_RPS)) // rot/s tolerance


@@ -93,12 +93,14 @@ private void configureBindings() {
 //        );

        swerve.setDefaultCommand(


…in v2 Part 1 — API style (team preference: constructors + instance methods, no static factories): - Robot code uses new Gains / MotionConstraints / CurrentBudget / MechanismConfig, direct StateMachine construction, and simRotationalModel/simArmModel config methods - SwerveConfig becomes pure public constants; RobotContainer constructs SwerveSubsystem directly and wires vision + AutoBuilder itself Part 2 — v1 parity (nothing lost from the old library): - SwerveSubsystem covers every v1 Swerve capability: robot-centric + heading-locked drive (turnToAngle), pathfindToPose / pathfindThenFollowPath (missing-file Alert kept), pathfind with driver override, reset heading/pose, coast, pose/heading/speeds getters, sawTagRecently trigger, Field2d, drivetrain SysId (translation/steer) - New utils: AllianceFlip.AllianceTranslation/AlliancePose, LedStrip (WPILib LEDPattern), LoopTimer, CAN-bus health in FaultReporter - Robot code switched off v1 utils (AllianceUtils/LEDs/ PerformanceMetricsTracker/LoggablePS5Controller/CANHealthMonitor); flywheel LED blink binding moved to RobotContainer via a new shooter.flywheelSpinning trigger; per-loop RT-priority-99 call dropped (audit P-15); orphaned LedState deleted - StateMachine schedules whileIn commands via CommandScheduler (Command.schedule() deprecated in 2026) - docs/V1_TO_V2_MAPPING.md: every v1 class -> ported / replaced / dropped Verified: build green, 9/9 tests, sim boots with no exceptions. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Self-contained HTML handbook covering the whole workflow: environment setup, project layout, the mental model, the five building-block value objects, a step-by-step first subsystem, every mechanism archetype with all config options, state machines, superstructure, swerve + vision + auto, telemetry/tuning/SysId, simulation, the robot loop, deploy, troubleshooting, checklists, and full config/API reference appendices. Features: sticky scrollspy nav, light/dark toggle, syntax-highlighted copy-paste code blocks, callout boxes, step lists, checklists, reading progress bar, print + mobile styles. No external dependencies. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Prescriptive, linear recipe book: one recipe per mechanism (arm/pivot/ hood, turret, elevator, flywheel, roller) and per subsystem, each as 'make these files -> paste this -> change these numbers -> call these functions'. Includes the exact state-machine build sequence, a complete worked subsystem, superstructure fan-out, swerve, and button/auto binding. Every archetype has a REQUIRED/OPTIONAL function table. Same interactive shell as the handbook (nav, dark mode, copy buttons). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

…enced devs) Decision-driven reference: each MechanismConfig field is posed as a question with branches that map straight to the exact call (feedback source, control mode, motion profile, gains, gravity type, current budget, limits/wrap, sim model). Conceptual and visual — inline SVG diagrams for the layer stack, setpoint data flow, archetype decision tree, motion-profile shapes (trapezoid vs Expo), gain-composition sum, current-limit split, state-transition graph, superstructure fan-out, and the loop timeline. Plus an archetype capability matrix and a 'decisions -> finished config' annotated example. Dark theme default. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Full Hebrew translation of the prose; all code, API names, and diagram labels stay English. Layout is direction-aware via CSS logical properties (border-inline, inset-inline, text-align:start) so it works in both RTL and LTR, with a direction toggle. Code blocks and inline identifiers are forced LTR + bidi-isolated so things like .fusedCANcoder(...) never scramble inside Hebrew text. Diagram labels kept English with Hebrew captions to avoid SVG bidi issues. Dark default. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

…y line In-depth annotated walkthrough of the actual migrated Intake.java: every config line, the follower/opposed reasoning, VOLTAGE-vs-FOC, the two speed clamps, arm-cosine gravity, boot seeding, the state machine (whileIn vs onEnter, run() requirements, transitionFromAny), the pump command, the public API's two request styles, and the mandatory periodic(). Plus a subsystem-map diagram, a 'what the library did for you' table, and a v1-vs-v2 contrast. Reasoning callouts throughout. Dark default, same interactive shell. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Single self-contained HTML manual merging and superseding the earlier handbook/cookbook/decision-map/anatomy drafts, with full depth: - 39 sections covering EVERY library class: CANDeviceId, ControlMode, Gains, MotionConstraints, CurrentBudget, Motor, DeviceConfigs, SignalHub, MechanismConfig (every field: signature, default, Phoenix mapping, failure mode), Mechanism base, Positional/Linear/Velocity/ Roller archetypes, MechanismSim, StateMachine (exact semantics), Superstructure, SwerveSubsystem (every command), LimelightMegaTag2, DriveToPose, Autos, Telemetry/DogLog, TunableNumber, FaultReporter, LoopTimer, AllianceFlip, Zones, LedStrip - §21: the real Intake dissected line by line with reasoning (per-line config table, whileIn-vs-onEnter, run() requirements, boot seed, pump composition, API audiences, the invisible work, v1 contrast) - Workflows: environment setup, repo/commands, boot & loop order, simulation, SysId->FOC tuning procedure, deploy & bring-up, troubleshooting matrix, checklists, glossary - Dual-audience: BASICS/DEEP badges; decision blocks mapping physical questions to exact calls; SVG diagrams (layers, dataflow, profiles, gain composition, state graph, fan-out, loop timeline); sidebar search (/), scrollspy, dark default, copy buttons API tables verified against the actual source (grep of all public methods), not from memory. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

EXCALIB_MANUAL.html is now the one canonical doc. Folded in the last pieces the drafts had that it lacked (archetype decision tree + capability matrix at the top of Part V; 'where the numbers come from' table in §13.15), then removed the superseded drafts: handbook, cookbook, decision map (EN), subsystem anatomy — all fully contained in the manual. The Hebrew decision-map edition is kept as a separate language deliverable. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

…rent) path §35 rewritten: what quasistatic/dynamic measure; the exact SysIdRoutine the library builds (config, SignalLogger state channel, why the device is the data logger, requirements); binding snippets for mechanisms and the swerve; run-session and hoot-analysis procedures; §35.5 the FOC path — why voltage results can't be reused, the volts-as-amps idiom, the new sysIdQuasistaticTorque/DynamicTorque routines and forceAmps, analyzer channel mapping, the two-edit FOC switch, and the note that Expo kV/kA stay volt-referenced. §14 API table updated with the torque routines. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

… approved The robot has run entirely on frc.excalib2 since the migration; v1 was kept in-tree only until removal was approved. It now is. - Delete the whole frc.excalib package (53 files: motor wrappers, mechanisms, swerve, slam/aurora, imu, math, utilities, commands). - Constants.java: drop the dead SwerveConstants inner class (replaced by SwerveConfig) and all v1 imports; drop now-unused imports. - Robot.java: drop PeriodicScheduler.run() and TalonFXMotor.refreshAll() (nothing registered on either anymore — all motors are excalib2 Motor managed by SignalHub). - Remove the vestigial .gitmodules (v1 was tracked in-repo, not a real submodule) and the stale MOTOR_LIBRARY_CHANGES.md (v1-era report). Verified: ./gradlew build green (compile + unit tests); simulateJava boots clean with no exceptions. No frc.excalib references remain in src. Monologue remains (separate vendordep, not part of this approval). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

v1 is gone: architecture notes now describe a single library; README tree drops the excalib/ line and fixes stale entries (Motor not ExcaTalonFX, drive commands on SwerveSubsystem, LinearExtension/LoopTimer/ LedStrip); manual §2 marks frc/excalib as removed; mapping doc flips the 'still on v1' section to a done status. Monologue noted as the last independent cleanup. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Mechanism fields widened to public final (Intake fourBar/roller, Shooter flywheel/hood/turret, Transport drum/transport) and Superstructure.transport exposed, so a characterization harness can reach them. RobotContainer.configureTestBindings(), gated to DriverStation Test mode via RobotModeTriggers.test() so nothing fires in teleop/auto: - L1: coast the whole robot (swerve + intake four-bar + turret) for zeroing - D-pad: SysId (quasistatic/dynamic, fwd/rev) on a one-line-selectable 'mechanism under test' — swap to the torque variants for FOC characterization - triangle/cross: intake OPEN/CLOSE direction+range check - circle: low fixed-speed flywheel spin (direction/readiness check) Teleop and autonomous wiring untouched. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

@nt

Removes the last legacy logging dependency: - Robot: drop Monologue.setupMonologue() and updateAll() - RobotContainer: drop 'implements Logged' + the @nt getter (its value wasn't load-bearing; live tuning is TunableNumber now) - BallCounter: drop 'implements Logged' + @Log.NT (the count is already logged to DogLog by the Shooter as Shooter/BallCount) - build.gradle: remove the com.github.shueja:Monologue vendordep Verified: build green; simulateJava boots clean and the WPILib DataLog is writing (logs/FRC_*.wpilog) alongside DogLog's NT/DataLog output and Phoenix SignalLogger. No monologue references remain anywhere. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

- docs/BRINGUP_TESTPLAN.html: on-robot first-run walk-through, ordered by risk. Pre-power (Tuner X), deploy & clean boot, proving DogLog + the .wpilog data log, the Test-mode harness, the per-mechanism recipe with exact DogLog keys and pass/fail, rollers/current-spike, the careful swerve-on-blocks section, vision, superstructure goals, auto dry-run, the SysId->FOC session, an if-it-fails matrix, and a key cheat-sheet. - CLAUDE.md / V1_TO_V2_MAPPING.md: Monologue retired (DogLog is the sole telemetry system); corrected the robot-loop description. - Manual doc index links the bring-up plan. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

YehudaRothstein and others added 6 commits July 2, 2026 18:37

docs: update README and CLAUDE.md for ExcaLib v2

3c504e1

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Copilot AI review requested due to automatic review settings July 2, 2026 15:57

Copilot started reviewing on behalf of YehudaRothstein July 2, 2026 15:57 View session

Copilot AI reviewed Jul 2, 2026

View reviewed changes

YehudaRothstein and others added 20 commits July 2, 2026 19:14

Merge branch 'feature/lib-rebuild' into feature/robot-migration

9aa3c5b

Merge branch 'feature/lib-rebuild' into feature/robot-migration

4f75f76

docs: update v1->v2 mapping for the ExcaTalonFX -> Motor rename

b6d25ce

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

style(robot): static-import NeutralModeValue.Coast in Intake and Shooter

be487f8

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Merge branch 'feature/lib-rebuild' into feature/robot-migration

289dabe

docs: mark LinearExtension as ported in the v1->v2 mapping

e5db77b

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Merge branch 'feature/lib-rebuild' into feature/robot-migration

1186bd9

YehudaRothstein deployed to github-pages July 5, 2026 18:52 — with GitHub Pages View deployment

Navigation Menu

Uh oh!

Robot-code migration onto ExcaLib v2 (stacked on #47) - #48

Robot-code migration onto ExcaLib v2 (stacked on #47)#48
YehudaRothstein wants to merge 27 commits into
feature/lib-rebuildfrom
feature/robot-migration

YehudaRothstein commented Jul 2, 2026 •

edited

Loading

Uh oh!

coderabbitai Bot commented Jul 2, 2026 •

edited

Loading

Review skipped

Uh oh!

Copilot AI left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

		@@ -93,12 +93,14 @@ private void configureBindings() {
		// );

		swerve.setDefaultCommand(

Uh oh!

Conversation

YehudaRothstein commented Jul 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What this is

Migration map

🟥 Behavior deltas — read before driving

Testing evidence

Follow-ups

Update: API style pass + full v1 feature parity

Update: ExcaLib v1 removed ✅ (approved)

Uh oh!

coderabbitai Bot commented Jul 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review skipped

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

YehudaRothstein commented Jul 2, 2026 •

edited

Loading

coderabbitai Bot commented Jul 2, 2026 •

edited

Loading