[BotshareAI]

Summary

Phase 1 of the 1 November 2026 alignment: a read-only audit of the 14 repositories in rollout/repositories.yaml against decisions.yaml on main 1678735 (in force P-03-rev20.2). It adds AUDIT-2026-11-01.md and updates STATE.md. No decision value changes.

The report contains:

• one table per repository (file, line, class, decision ID or UC1-nn, current and correct text, source, whether it is fixable, and overlap with open PRs);
• counts per class and repository;
• the source conflicts S1 to S17 (register vs P-03 rev20.2 vs Use Case 1 baseline);
• the full list of class C, D and E findings for the owner.

Totals: 117 checker lines (classified per repository); 352 reading-review rows: A 97, B 120, C 23, D 39, E 73.

Phase 2 fix PRs (class A and B only, all draft). Each repository section of the report lists the rows its PR fixed.

• docs: current configuration ID example, 2.0 LiDAR and base controller in the roadmap, profile wording (P-03 rev20.2) #80
• docs: label existing-robot (Gate A) material, name the 2.0 parts, docking is positioning only (P-03 rev20.2) openamr-platform-sw#107
• docs: label Gate A (Teensy) material and legacy compute; near-field sensor supplies (P-03 rev20.2) openamr-platform-fw#32
• docs: label existing-robot hardware legacy and state the 2.0 baseline values (P-03 rev20.2, BOM Issue 7.14) openamr-platform-hw#42
• docs: align with P-03 rev20.2 (head camera hp30, lift node CTL-017, holding-test gate) openamr-upperbody-fw#13
• docs: align with P-03 rev20.2 (head-camera lens x, rotatable tablet mount, lift data, holding-test gate) openamr-upperbody-hw#14
• docs: 2.0 charging path in the nav_msgs README; correct stale STATE.md facts openamrobot-interfaces#32
• docs: combined-model inventory follows the current lift register (holding-test gate, CTL-017, BOM 7.14) openamrobot-manipulation#26
• docs: Stop (software) wording in docs and AGENTS.md; charging station card is not charge state openamrobot-ui#58
• docs: align the 2.0 reference with P-03 rev20.2 (BOM Issue 7.14, sensor poses, Stop (software), legacy labels) and test it openamrobot-docs#67
• docs: what sllidar_ros2 and micro_ros_agent are for; PR #11 pin already on main openamrobot-manifest#20
• docs: v0.0.1 compute labelled legacy; citation says the lift and arms are 2.0 openamrobot-release#17
• openamr-upperbody-sw: audit only, no PR. openamrobot-comm: no class A or B finding.

Class C, D and E findings stay open as owner decisions.

Work package

Repository alignment with P-03 rev20.2 (item 26, change request #77) and the Use Case 1 baseline rev18.5; no separate issue.

Integration Gate

• Open PRs in every repository were fetched (refs/pull/<N>/head) and diffed; each finding records "covered by #N", "same file, different lines" or none.
• In this repository, decisions: PLATFORM-PIN moves to platform-sw 131e500 (2.0 base) #74 (PLATFORM-PIN), harness: autonomous improvement loop and decisions.yaml lifecycle (design, off until founder approval on #43) #70, watchdog: daily software-health report (CI, open PRs, rc.1 issues) and openamrobot-embodied-ai in the scan #69, fix(harness): enforce publish guard against unsafe commit history and tier bypass #76, ci: add baseline ROS 2 CI image #41 and docs: current configuration ID example, 2.0 LiDAR and base controller in the roadmap, profile wording (P-03 rev20.2) #80 do not touch the two files changed here.
• Reused: tools/check_decisions.py (unchanged), the register's fix hints and its WATCHDOG wording.
• Rejected: adding the report to the register exclude list. That would weaken a check, so the 15 lines that quote retired text carry decision-allow markers instead.

Decisions impact

Consumer only. The report cites every recorded decision ID. It proposes register changes as owner decisions (S1 to S17: IMU-TOPIC-OWNERSHIP, precharge, chest E-stop wording, K1/K2, NEAR-FIELD-FRAMES, PLATFORM-PIN, item 24 equipment, UI-GATE-A naming) and makes none.

Tests

No behaviour change and no test change. Repository suite run on the head: bash rollout/verify.sh: 289 of 289 tests executed, 0 skipped, PASS.

Evidence

Base SHA: 1678735
Head SHA: 6b0376e

bash rollout/verify.sh                       # Tests executed: 289 of 289 (skipped 0); PASS: all detected verification stages
python3 tools/check_decisions.py --decisions decisions.yaml --maintainers maintainers.yaml --validate-only   # exit 0; 46 loaded, 41 recorded, 2 open, 3 superseded
python3 tools/check_agent_rules.py --canonical agent-rules/SHARED_RULES.md --file AGENTS.md                  # PASS AGENTS.md (v2)
python3 tools/check_decisions.py --decisions decisions.yaml --maintainers maintainers.yaml --root . --repository .github --changed-files changed.txt   # 0 contradiction(s), 15 allowed, 2 changed files
python3 tools/check_public_extract.py --root . --repository .github --changed-files changed.txt                # 0 finding(s)
# audit scan, once per repository (14):
python3 tools/check_decisions.py --decisions decisions.yaml --maintainers maintainers.yaml --root <checkout> --repository <name> --json <name>.json

Dependencies

None.

Safety impact

None. Text only; no safety function is changed. Safety-path findings (E-stop, contactor, precharge, watchdog wording and code) are reported for the platform lead and are not edited.

STATE.md

Updated: #78 merged as 1678735, in_force on main, last verify PASS on 1678735, an audit row, and the Not verified line for BOM Issue 7.14 content.

Not verified

• BOM Issue 7.14 content: only its metadata was read. Finding S14 is a reviewer report and was not checked by the lead.
• P-03 rev20.2 and the Use Case 1 baseline were read once on 2026-11-01. Later edits to them are not reflected.
• The reading review was done with assistance. Each table row is a claim for the owning reviewer to confirm. Line numbers refer to the audited SHAs.
• openamrobot-embodied-ai is not in rollout/repositories.yaml. It was scanned (0 findings) but is not counted.
• No build, simulation or hardware evidence: this audit proves textual consistency only.

AI disclosure

Claude Code produced the audit: it ran the register scans, did the reading review with parallel read-only subagents, assembled the report, edited STATE.md and prepared the fix PRs. Personal names, prices and contact data found in repositories were replaced in the report, and lines quoting retired text were marked. The work-package owner should write adopt, adapt or reject in this thread.

Contribution terms

• Every commit is signed off under the DCO.
• I am covered by an accepted Contributor Agreement.
• Third-party material is identified with source and licence.
• No confidential, personal, credential or export-controlled information is included.
• No partner, customer or private person is named; the application is Use_Case_1.

🤖 Generated with Claude Code

https://claude.ai/code/session_01DZnPENAVyrfCiDEPBa8zK7