#539 - #534 C: No Project.load outside open — four call sites - jared/cc-os - Forgejo: Beyond coding. We forge. | Jared Swanson

jared commented

2026-09-08 20:15:05 +00:00

Owner

Context

Child of #534 (ADR-0169). Parent design points and comments on #534. Lands after #535.

Tasks

runner/brief_facts.rb:134: replace Project.load(@project_root, target: nil) with @implementation.project(@project_root) (BriefFacts already receives the implementation from start_subagent.rb:43)
runner/gate_commands.rb:29: GateHandoff (gate_handoff.rb:89) already holds the implementation; pass implementation.project(project_root) (or the implementation) into GateCommands and delete GateCommands#project's own Project.load
runner/tea_tracker.rb:146: keep Project.load(@root, target: nil) as a documented exemption; it runs from OpenImplementation#initialize before any implementation row exists and only reads tracker keys, which are target-agnostic. Add a one-line comment citing ADR-0169 and #539
regression test plugins/os-sdlc/tests/runner_project_load_guard_test.rb mirroring settlement_vocabulary_guard_test.rb: glob lib/os_sdlc/runner/**/*.rb, fail on any Project.load( outside an exemption list of open_implementation.rb and tea_tracker.rb, reporting path:line

Acceptance criteria

grep -rn "Project.load(" plugins/os-sdlc/lib/os_sdlc/runner returns only open_implementation.rb and tea_tracker.rb
the guard test is green and fails when a Project.load is added to any other runner file
os-sdlc suite green

Blocking edges

Blocked by: #535 and #538

Origin

• Trigger: Agent-initiated from ticket-skeptic gate (CREATE verdict)
• Improvised this session: none
• Chain: DESIGN ← ADR-0169
• Root candidate: #534
• Where: n/a
• Session: 1a3b7fd0-0319-465b-8414-0ab70560de3d
• Transcript: (from session context)
• Judge: claude-sonnet-5 on 2026-09-08; user-decided reframe recorded in ADR-0169; former child G merged into B

## Context Child of #534 (ADR-0169). Parent design points and comments on #534. Lands after #535. ## Tasks - [ ] runner/brief_facts.rb:134: replace `Project.load(@project_root, target: nil)` with `@implementation.project(@project_root)` (BriefFacts already receives the implementation from start_subagent.rb:43) - [ ] runner/gate_commands.rb:29: GateHandoff (gate_handoff.rb:89) already holds the implementation; pass `implementation.project(project_root)` (or the implementation) into GateCommands and delete GateCommands#project's own Project.load - [ ] runner/tea_tracker.rb:146: keep `Project.load(@root, target: nil)` as a documented exemption; it runs from OpenImplementation#initialize before any implementation row exists and only reads tracker keys, which are target-agnostic. Add a one-line comment citing ADR-0169 and #539 - [ ] regression test `plugins/os-sdlc/tests/runner_project_load_guard_test.rb` mirroring settlement_vocabulary_guard_test.rb: glob lib/os_sdlc/runner/**/*.rb, fail on any `Project.load(` outside an exemption list of open_implementation.rb and tea_tracker.rb, reporting path:line ## Acceptance criteria - [ ] grep -rn "Project.load(" plugins/os-sdlc/lib/os_sdlc/runner returns only open_implementation.rb and tea_tracker.rb - [ ] the guard test is green and fails when a Project.load is added to any other runner file - [ ] os-sdlc suite green ## Blocking edges Blocked by: #535 and #538 ## Origin • Trigger: Agent-initiated from ticket-skeptic gate (CREATE verdict) • Improvised this session: none • Chain: DESIGN ← ADR-0169 • Root candidate: #534 • Where: n/a • Session: 1a3b7fd0-0319-465b-8414-0ab70560de3d • Transcript: (from session context) • Judge: claude-sonnet-5 on 2026-09-08; user-decided reframe recorded in ADR-0169; former child G merged into B

jared added the

P1

plugin/os-sdlc

filed-by/agent

labels

2026-09-08 20:15:05 +00:00

jared referenced this issue

2026-09-08 20:15:15 +00:00

#534 D: Fact owners on the models #540

jared referenced this issue

2026-09-08 20:15:40 +00:00

Brief reads every fact from the Implementation aggregate; project config binds at open (was: brief ignores target) #534

jared referenced this issue

2026-09-08 20:48:02 +00:00

Diff-test gate loops when a lib file has only scenario-split tests and no mirrored _test.rb #535

jared referenced this issue

2026-09-09 14:29:34 +00:00

#534 A: Project: target keyword required; JSON round-trip #537

jared referenced this issue

2026-09-09 15:06:58 +00:00

#534 B: Schema 21: project_config column; bind at open; Implementation#project #538

jared commented

2026-09-09 15:09:53 +00:00

Author

Owner

Second code-probe done 2026-09-09 (claude-sonnet-5, session d233a2f7-977a-419d-99ea-011b6e1e66f8). Findings: brief_facts.rb and gate_commands.rb both already have an Implementation in their caller chain (start_subagent.rb:43, gate_handoff.rb:89) and thread implementation.project(root) in. tea_tracker.rb:146 is reached from OpenImplementation#initialize before any row exists and is target-agnostic; kept as a documented exemption. Five further Project.load sites sit outside runner/ (autocorrect_prepass.rb, lint_command_support.rb, lint_worklist_target.rb, project_config_command.rb, scope_policy.rb); out of this ticket's stated scope, captured separately. Tasks rewritten; ticket is ready.

jared added the

ready-for-agent

label

2026-09-09 15:10:02 +00:00

jared referenced this issue

2026-09-09 15:10:37 +00:00

Project.load sites outside runner/ after ADR-0169 #545

jared commented

2026-09-09 15:10:47 +00:00

Author

Owner

Work started via /os-sdlc:implement on branch ticket-539 (session d233a2f7-977a-419d-99ea-011b6e1e66f8). Deviation carried from #537/#538: main's brief_facts.rb hot-patched with target: @implementation.target for the run (#534 hook gap); restored at finish.

jared commented

2026-09-09 15:43:25 +00:00

Author

Owner

Captured map escalate-route issue (#539's increment 2): unexpected_pass on later increments is a legitimate "already done" signal (refactorer generalises scope), not a defect. Deferred: route unexpected_pass:non-first-increment to behavior-verifier or allow test-writer guard/characterisation declaration so diff-test expects green. Skeptic verdict: depth-1 anti-loop rule (parent is AI-created follow-up).

jared commented

2026-09-09 15:44:43 +00:00

Author

Owner

Resolution

Done: Under lib/os_sdlc/runner only open_implementation.rb and tea_tracker.rb call Project.load. Project::Source plays the project-source role for collaborators without an implementation; an implementation row plays it through #project. BriefFacts and GateCommands (fed the implementation by GateHandoff) ask their source, so brief facts and gate commands come from the config bound at open. tea_tracker.rb keeps a one-line documented exemption (pre-open, target-agnostic). runner_project_load_guard_test.rb fails on any other Project.load under runner/, proven by emptying its exemption list.

Evidence: Pipeline /os-sdlc:implement 539, implementation 1 on the schema-21 db, session d233a2f7-977a-419d-99ea-011b6e1e66f8, dispatches 1-15. Commits afb5440 (refactor) and b9eb84d (history row) on branch ticket-539, merged to main 2026-09-09. Suite 1097 runs green. Deviations: (1) the run ended in escalation unexpected_pass:test-writer/diff-test at dispatch 15: increment 1's refactorer had already rewired every runner site, so increment 2's two tests (gate_commands_implementation_project_test.rb, brief_facts_project_source_test.rb) passed on creation; they were kept as they are relevant and public. (2) The guard test is green by construction and cannot enter red-first, so it was written by hand as test-only work (sonnet), outside the pipeline gates, and is disclosed here. (3) The escalation procedure asks for one capture ticket; the ticket-skeptic dropped it under the depth-1 anti-loop rule because #539 is agent-created, and the cause is recorded as a comment on this ticket instead. (4) main's brief_facts.rb hot-patched for the run (#534 hook gap), restored at finish.

Follow-ups: #545 (Project.load sites outside runner/); #541 (BriefFacts asks, never queries) and #542 next in the #534 chain; the unexpected_pass-on-later-increment map gap is recorded in this ticket's comments, not ticketed, per the skeptic's DROP verdict

## Resolution **Done:** Under lib/os_sdlc/runner only open_implementation.rb and tea_tracker.rb call Project.load. Project::Source plays the project-source role for collaborators without an implementation; an implementation row plays it through #project. BriefFacts and GateCommands (fed the implementation by GateHandoff) ask their source, so brief facts and gate commands come from the config bound at open. tea_tracker.rb keeps a one-line documented exemption (pre-open, target-agnostic). runner_project_load_guard_test.rb fails on any other Project.load under runner/, proven by emptying its exemption list. **Evidence:** Pipeline /os-sdlc:implement 539, implementation 1 on the schema-21 db, session d233a2f7-977a-419d-99ea-011b6e1e66f8, dispatches 1-15. Commits afb5440 (refactor) and b9eb84d (history row) on branch ticket-539, merged to main 2026-09-09. Suite 1097 runs green. Deviations: (1) the run ended in escalation `unexpected_pass:test-writer/diff-test` at dispatch 15: increment 1's refactorer had already rewired every runner site, so increment 2's two tests (gate_commands_implementation_project_test.rb, brief_facts_project_source_test.rb) passed on creation; they were kept as they are relevant and public. (2) The guard test is green by construction and cannot enter red-first, so it was written by hand as test-only work (sonnet), outside the pipeline gates, and is disclosed here. (3) The escalation procedure asks for one capture ticket; the ticket-skeptic dropped it under the depth-1 anti-loop rule because #539 is agent-created, and the cause is recorded as a comment on this ticket instead. (4) main's brief_facts.rb hot-patched for the run (#534 hook gap), restored at finish. **Follow-ups:** #545 (Project.load sites outside runner/); #541 (BriefFacts asks, never queries) and #542 next in the #534 chain; the unexpected_pass-on-later-increment map gap is recorded in this ticket's comments, not ticketed, per the skeptic's DROP verdict

jared closed this issue

2026-09-09 15:44:44 +00:00

jared referenced this issue

2026-09-09 18:54:56 +00:00

Brief reads every fact from the Implementation aggregate; project config binds at open (was: brief ignores target) #534

jared referenced this issue from a commit

2026-09-09 19:05:19 +00:00

feat(os-sdlc): Project.load requires target:, JSON round-trip without root (#537)

jared referenced this issue from a commit

2026-09-09 19:05:19 +00:00

refactor(os-sdlc): runner reads the bound project, never Project.load (#539)

jared referenced this issue from a commit