eda: wire the playbook coordinator into production task intake #447

Closed
opened 2026-10-27 13:36:22 +00:00 by sam · 0 comments
Owner

Found by a read-only decision lane on 2026-09-18 while resolving .local/issue-tracker
findings eda-engine-1 and eda-engine-2.

The coordinator is built but nothing dispatches through it:

  • internal/domain/eda/playbook/playbook.go:262 Store.Dispatch and :230 Playbook.Match
    have no production caller. bootstrap.go:780 calls playbook.Load, never Dispatch.
  • internal/domain/eda/module/module.go:198 ModuleRegistry.Invoke and :169 decodeLogArgs
    have no production caller. bootstrap.go:749 constructs module.New(), nothing invokes.
  • rg '\.Invoke\(' -g '!*_test.go' internal/ returns nothing.
  • task reachability reports Playbook.Match, Store.Dispatch and Env.Eval unreachable.

docs/prds/eda-playbook-engine.md section "Open questions -> 4. First implementation slice"
states plainly: "The coordinator is NOT yet wired into production task intake
(daemon.go/pollNATS/publishTask) -- that is a follow-up ticket". That ticket does not
exist: bd search coordinator returns only t2db.15, which shipped the coordinator and
is closed. The epic's open children (t2db.19, t2db.20) are unrelated.

This bead is that missing ticket. Without it the wiring is untracked, and the
half-built coordinator looks like dead code to every future audit -- which is exactly
how it was filed twice as a P1 defect.

Scope note: this is a cutover of the intake path, not a small fix. It must not delete
the coordinator (it is the ratified epic's core); it wires it.

Acceptance criteria

A forge-triggered task in production flows through playbook.Store.Dispatch to module.ModuleRegistry.Invoke rather than only workflow.Route(); the wiring lives in the daemon's intake path (daemon.go pollNATS/publishTask) and has a test that fails when the coordinator is bypassed.

Close reason

Wired at the daemon's definition pin (Daemon.resolveWorkflowID), not pollNATS/publishTask: routing moved out of the agent process when workflow definitions became DB rows pinned before dispatch. Commit 61887fd7. The AC's module.ModuleRegistry.Invoke half is not reachable under the t2db.15 hard boundary and needs the playbook_dispatches ledger first -> archie-core-t2db.24.


Imported from beads archie-core-t2db.23 · GitHub #911 · type task · priority P1 · created 2026-09-18 · closed 2026-09-21

Found by a read-only decision lane on 2026-09-18 while resolving .local/issue-tracker findings eda-engine-1 and eda-engine-2. The coordinator is built but nothing dispatches through it: - internal/domain/eda/playbook/playbook.go:262 Store.Dispatch and :230 Playbook.Match have no production caller. bootstrap.go:780 calls playbook.Load, never Dispatch. - internal/domain/eda/module/module.go:198 ModuleRegistry.Invoke and :169 decodeLogArgs have no production caller. bootstrap.go:749 constructs module.New(), nothing invokes. - `rg '\.Invoke\(' -g '!*_test.go' internal/` returns nothing. - task reachability reports Playbook.Match, Store.Dispatch and Env.Eval unreachable. docs/prds/eda-playbook-engine.md section "Open questions -> 4. First implementation slice" states plainly: "The coordinator is NOT yet wired into production task intake (daemon.go/pollNATS/publishTask) -- that is a follow-up ticket". That ticket does not exist: `bd search coordinator` returns only t2db.15, which shipped the coordinator and is closed. The epic's open children (t2db.19, t2db.20) are unrelated. This bead is that missing ticket. Without it the wiring is untracked, and the half-built coordinator looks like dead code to every future audit -- which is exactly how it was filed twice as a P1 defect. Scope note: this is a cutover of the intake path, not a small fix. It must not delete the coordinator (it is the ratified epic's core); it wires it. ### Acceptance criteria A forge-triggered task in production flows through playbook.Store.Dispatch to module.ModuleRegistry.Invoke rather than only workflow.Route(); the wiring lives in the daemon's intake path (daemon.go pollNATS/publishTask) and has a test that fails when the coordinator is bypassed. ### Close reason Wired at the daemon's definition pin (Daemon.resolveWorkflowID), not pollNATS/publishTask: routing moved out of the agent process when workflow definitions became DB rows pinned before dispatch. Commit 61887fd7. The AC's module.ModuleRegistry.Invoke half is not reachable under the t2db.15 hard boundary and needs the playbook_dispatches ledger first -> archie-core-t2db.24. --- _Imported from beads `archie-core-t2db.23` · GitHub #911 · type task · priority P1 · created 2026-09-18 · closed 2026-09-21_
sam added the hightask labels 2026-10-27 13:36:22 +00:00
sam self-assigned this 2026-10-27 13:36:22 +00:00
sam closed this issue 2026-10-27 13:36:22 +00:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Reference: sam/archie#447