Plan 321 P0 — record the shepherd's four measured inputs

[munchist]

Records the four measured inputs of the shepherd's P0 in docs/plans/321-the-loop-recovers-from-ci.md §5, each with the command that produced it, and sets P0's exit as met — then, on two review rounds, changes the decisions the measurements contradict (a measurement that contradicts a decision changes the decision, never sits beside it):

1. Schedule-to-skill-input path (code read at a7939a7) — a schedule starts an agent run, never a skill; the skill runs because that agent calls workflow.run.start with the inputs its template carries. D1/D2 name the design the code already allows: the agent that decides nothing is a custom agent — an operator-authored step-graph agent (agent.manage.create, packages/schemas/src/runtime/agentTarget.ts custom-agent) whose definition is one operation step, workflow.run.start of local-pr-shepherd with bindingId as a literal, and no model turn at all — no ai.agent.turn, no prompt, nothing to drift. The schedule targets that agent by slug; the Local PR Shepherd bundle installs the agent beside the skill so the operator creates nothing by hand; D2 names the same agent. Templates carry static JSON, $now, and (on completion) the source runId/status/outputRef; $memoryRef always fails the firing; the template is checked only for non-emptiness at creation; a repo coordinate is campaign config, not a run input.
2. GitHub rate budget (gh api rate_limit live probe of the binding's token) — core limit 5,000 requests/hour; x-ratelimit-*/Retry-After reach the agent unparsed; the generic retry never reads them, the exhausted-limit 403 is not retried, POST/PATCH calls are never retried on 429. The interval is derived in the plan: one pass for N open PRs costs the list + one check-runs read per PR + one jobs read per failing run + one log read per failing job, and the default interval follows with a wide margin, written as the number and the arithmetic.
3. Size of a failing job's log (gh api repos/aflowai/aflow/actions/runs/37460214524/attempts/1/jobs, then gh api repos/aflowai/aflow/actions/jobs/<job_id>/logs) — Test job 923,330 bytes / 5,955 lines (failure in the last 280), Dependency audit job 32,310 bytes / 421 lines (advisories in the last 35). A last-300-lines, under-60 KB tail bound carried the failure for both job kinds observed; a bound needs no endpoint binding to hold.
4. Attention list (workflow.run.list_attention) — items name a run, never a branch, PR or repo. D3.4 names the mechanism that replaces it, a bounded read of the space's runs by workflow and input field, and §4 turns its conditional into the packages/schemas change that read needs, with the P2 exit following.

Status: P0 measured, P1 next. Citations are repository-relative paths (the P0.1 $memoryRef line range names its own file); no systemRole: 'helmsman' remains; nothing else in the document changes.