Do you think it's time that we review our prompts/dispatch packets/cha

You are being dispatched from flower Brief #97: Do you think it's time that we review our prompts/dispatch packets/cha

Recall pointer:
- Use recall_brief with id 97 for the full folder if you need provenance.

Target:
- project: flower (/Users/mikeferrara/Documents/code/flower)
- branch: flower/97-charter-checkin-fix
- worktree: not specified
- kind: fresh

Current brief spec:
## Ask
Audit and update the daemon **charters / dispatch packets / shared conventions** so they match current
procedure — specifically the **`flower:daemon-checkin` command** (Brief #67) that resolves identity
(solo_process_id, session_id, context) from the environment. Also decide how
`~/Documents/code/_solo-playbook.md` should relate to these prompts.

## Evidence (audited 2026-07-03 via recall_charters — answers the operator's direct question)
**The charters were NOT updated for the checkin command. They still mandate the bare `daemon_checkin` MCP tool:**
- **Orchestrator charter** (`daemon_charter.orchestrator.default` v1): heartbeat step says
  *"daemon_checkin(role, project) including solo_process_id, session_id, context_size, and usage"* — i.e.
  the bare MCP tool **with hand-passed ids**. Never mentions the command.
- **Shared `daemon_conventions` block** (used by ops + refine charters): offers *"the check-in command
  (`/flower-checkin` / `php artisan flower:daemon-checkin`) **or** daemon_checkin with role, project,
  current_cadence, actor_ref"* as **equivalent** — but the bare-tool variant with no ids does NOT
  auto-resolve identity, so it writes null solo_process_id/session_id.
- **Internal contradiction:** the conventions say *"do NOT hand-copy ids"* while the orchestrator charter
  says *"include solo_process_id, session_id"* — opposite instructions.

**Real consequence, observed on this refine daemon (2026-07-03):** the spawn-packet kickoff told the
refine daemon to call bare `daemon_checkin` with only role/project/actor_ref → ids came through **null**;
then the `flower:daemon-checkin` command created a **second, separate roster row** (duplicate: stale
null-id row #8 + live correct row #9) because `daemon_register_expected` placeholders and command-checkins
don't reconcile. (Filed as flower_feedback / mcp_issue.)

## Scope
1. **Rewrite the heartbeat instruction everywhere to a single path:** the **`flower:daemon-checkin`
   command** (or `/flower-checkin <role> <actor-ref>`) is THE heartbeat mechanism. It auto-resolves
   solo_process_id + session_id + context from env; agents must NOT hand-pass or self-estimate those.
   - Fix the orchestrator charter's "daemon_checkin(...) including solo_process_id, session_id" line.
   - Fix the shared `daemon_conventions` block: demote the bare `daemon_checkin` MCP tool from a
     heartbeat option (keep it only for non-heartbeat daemon ops). Remove the "or daemon_checkin ...
     actor_ref" equivalence that creates the trap.
   - Fix the **daemon spawn-packet kickoff** (the render packet + `daemon_charter.<role>.default`) so the
     step-4 kickoff says "run `flower:daemon-checkin`", not "call daemon_checkin".
2. **Reconciliation fix (ties to the duplicate-row bug):** the command should ADOPT an existing
   `expected` placeholder for the same (role, project) instead of creating a new row — or
   register_expected should be keyed so the command updates it. May split to its own code brief off the
   feedback; noted here as root cause of the observed duplicate.
3. **_solo-playbook.md relationship (operator's question) — recommendation:** charters should
   **link/reference** the playbook as the deep operator reference, NOT duplicate it. Agents spawned in a
   project already get CLAUDE.md/AGENTS.md routing to `_solo-playbook.md` / `_flower-playbook.md`, so the
   charters need only the flower-specific daemon loop + the correct heartbeat command + a pointer.
   (Open to override if you'd rather inline more.)

## Acceptance
- All three daemon charters + the shared conventions block + the spawn-packet kickoff use the
  `flower:daemon-checkin` command as the sole heartbeat path; no charter instructs hand-passing ids.
- The contradiction (do-not-hand-copy vs include-ids) is gone.
- Duplicate-row reconciliation is fixed or a dedicated code brief is filed for it.
- `_solo-playbook.md` relationship documented (link, not duplicate — or per operator).
- `php artisan test` green + pint. `Brief: #97` trailer.

## Provenance
Operator note (#97, 2026-07-03). Refined by flower-refine (2026-07-03) with a live charter audit
(recall_charters) + the observed duplicate-row bug as motivating evidence. Related: Brief #67 (the
checkin command), the flower_feedback mcp_issue on the duplicate row.

Recent/key trace events:
[1] participant_joined operator:mike: (no body)
[2] note_added operator:mike: Do you think it's time that we review our prompts/dispatch packets/charters to daemons and agents to make sure they're up to date with recent/any changes to procedure or tools/calls/etc? I'm wondering if our orchestrator charters have been updated to handle/support the new checkin command? (or did it not work for some reason?)
[3] status_change operator:mike: (no body)
[4] spec_snapshot operator:mike: Do you think it's time that we review our prompts/dispatch packets/charters to daemons and agents to make sure they're up to date with recent/any changes to procedure or tools/calls/etc? I'm wondering if our orchestrator charters have been updated to handle/support the new checkin command? (or did it not work for some reason?)
[5] refinement operator:mike: Do you think it's time that we review our prompts/dispatch packets/charters to daemons and agents to make sure they're up to date with recent/any changes to procedure or tools/calls/etc? I'm wondering if our orchestrator charters have been updated to handle/support the new checkin command? (or did it not work for some reason?)

To go with this - should we review our ~/Documents/code/_solo-playbook.md and see what/how we need to integrate things into the prompts/charters/packets (or at all? because they still get it when they're spawned in those projects?)?
[6] participant_joined flower-refine: (no body)
[7] spec_snapshot flower-refine: Do you think it's time that we review our prompts/dispatch packets/charters to daemons and agents to make sure they're up to date with recent/any changes to procedure or tools/calls/etc? I'm wondering if our orchestrator charters have been updated to handle/support the new checkin command? (or did it not work for some reason?)

To go with this - should we review our ~/Documents/code/_solo-playbook.md and see what/how we need to integrate things into the prompts/charters/packets (or at all? because they still get it when they're spawned in those projects?)?
[8] refinement flower-refine: ## Ask
Audit and update the daemon **charters / dispatch packets / shared conventions** so they match current
procedure — specifically the **`flower:daemon-checkin` command** (Brief #67) that resolves identity
(solo_process_id, session_id, context) from the environment. Also decide how
`~/Documents/code/_solo-playbook.md` should relate to these prompts.

## Evidence (audited 2026-07-03 via recall_charters — answers the operator's direct question)
**The charters were NOT updated for the checkin command. They still mandate the bare `daemon_checkin` MCP tool:**
- **Orchestrator charter** (`daemon_charter.orchestrator.default` v1): heartbeat step says
  *"daemon_checkin(role, project) including solo_process_id, session_id, context_size, and usage"* — i.e.
  the bare MCP tool **with hand-passed ids**. Never mentions the command.
- **Shared `daemon_conventions` block** (used by ops + refine charters): offers *"the check-in command
  (`/flower-checkin` / `php artisan flower:daemon-checkin`) **or** daemon_checkin with role, project,
  current_cadence, actor_ref"* as **equivalent** — but the bare-tool variant with no ids does NOT
  auto-resolve identity, so it writes null solo_process_id/session_id.
- **Internal contradiction:** the conventions say *"do NOT hand-copy ids"* while the orchestrator charter
  says *"include solo_process_id, session_id"* — opposite instructions.

**Real consequence, observed on this refine daemon (2026-07-03):** the spawn-packet kickoff told the
refine daemon to call bare `daemon_checkin` with only role/project/actor_ref → ids came through **null**;
then the `flower:daemon-checkin` command created a **second, separate roster row** (duplicate: stale
null-id row #8 + live correct row #9) because `daemon_register_expected` placeholders and command-checkins
don't reconcile. (Filed as flower_feedback / mcp_issue.)

## Scope
1. **Rewrite the heartbeat instruction everywhere to a single path:** the **`flower:daemon-checkin`
   command** (or `/flower-checkin <role> <actor-ref>`) is THE heartbeat mechanism. It auto-resolves
   solo_process_id + session_id + context from env; agents must NOT hand-pass or self-estimate those.
   - Fix the orchestrator charter's "daemon_checkin(...) including solo_process_id, session_id" line.
   - Fix the shared `daemon_conventions` block: demote the bare `daemon_checkin` MCP tool from a
     heartbeat option (keep it only for non-heartbeat daemon ops). Remove the "or daemon_checkin ...
     actor_ref" equivalence that creates the trap.
   - Fix the **daemon spawn-packet kickoff** (the render packet + `daemon_charter.<role>.default`) so the
     step-4 kickoff says "run `flower:daemon-checkin`", not "call daemon_checkin".
2. **Reconciliation fix (ties to the duplicate-row bug):** the command should ADOPT an existing
   `expected` placeholder for the same (role, project) instead of creating a new row — or
   register_expected should be keyed so the command updates it. May split to its own code brief off the
   feedback; noted here as root cause of the observed duplicate.
3. **_solo-playbook.md relationship (operator's question) — recommendation:** charters should
   **link/reference** the playbook as the deep operator reference, NOT duplicate it. Agents spawned in a
   project already get CLAUDE.md/AGENTS.md routing to `_solo-playbook.md` / `_flower-playbook.md`, so the
   charters need only the flower-specific daemon loop + the correct heartbeat command + a pointer.
   (Open to override if you'd rather inline more.)

## Acceptance
- All three daemon charters + the shared conventions block + the spawn-packet kickoff use the
  `flower:daemon-checkin` command as the sole heartbeat path; no charter instructs hand-passing ids.
- The contradiction (do-not-hand-copy vs include-ids) is gone.
- Duplicate-row reconciliation is fixed or a dedicated code brief is filed for it.
- `_solo-playbook.md` relationship documented (link, not duplicate — or per operator).
- `php artisan test` green + pint. `Brief: #97` trailer.

## Provenance
Operator note (#97, 2026-07-03). Refined by flower-refine (2026-07-03) with a live charter audit
(recall_charters) + the observed duplicate-row bug as motivating evidence. Related: Brief #67 (the
checkin command), the flower_feedback mcp_issue on the duplicate row.
[9] status_change flower-refine: (no body)

Recommended linked context:
{
    "todos": [],
    "scratchpads": []
}

Execution notes:
- Treat the brief as the source of truth.
- Keep work scoped to this dispatch request.
- Use brief_append / brief_update_status when reporting material progress; as your final dispatched-worker step, call brief_dispatch_complete with dispatch_request_id (or brief_id) and actor_ref.
- Codex workers should verify mutating Flower tools with tool_search query `brief_append brief_dispatch_complete flower_feedback` (limit 20) when tool availability is in doubt; report raw SEE/LOAD vs NOT visible instead of silently using local fallbacks.
- Add a git commit trailer `Brief: #97` to every commit for this brief so flower can exact-link commits back to the brief.