Sean-Michael Night Shift
Three modes for unattended autonomous work.
| Mode | Who drives | What it does |
|---|---|---|
| PLAN | You + Claude | Decompose work into executable task list |
| RUN | Headless agents | Execute safe tasks → DIGEST |
| AUDIT LOOP | Headless agents | Audit → score → implement → repeat |
Risk Model
Applies to all modes. Agents only execute safe tasks unless explicitly unlocked.
| Tag | Agent behavior |
|---|---|
safe |
Execute — code changes, tests, docs, research |
review |
Execute, flag loudly in DIGEST |
risky |
Hard skip unless NIGHT_SHIFT_UNLOCK_RISKY=true |
Absolute hard stops — never execute regardless of unlock:
terraform apply/plan,kubectl apply, any infra provisioning- Database migrations against live instances
git pushto main/master- Anything not reversible with
git revert
Mode 1: PLAN
Invoke: > Night shift planning: [what you want done]
Interview the user for: repos in scope, done criteria, risk tags, dependencies.
Create tasks via TaskCreate. Write NIGHT_SHIFT_MANIFEST.yaml to repo root.
Task schema:
title short imperative phrase
WHAT exactly what to do
WHY what it unblocks
FILES key files likely involved
DONE verifiable exit criterion — not "looks good"
NOTES gotchas, prior art
risk safe | review | risky
deps [task-id, ...]
Rules: ≤3h per task · one concern per task · no DONE = not ready to run.
Before closing: TaskList — confirm no risky tasks are accidentally pending.
See references/manifest-schema.yaml for full MANIFEST fields including audit loop config.
Mode 2: RUN
export CLAUDE_CODE_TASK_LIST_ID="$(grep task_list_id NIGHT_SHIFT_MANIFEST.yaml | awk '{print $2}')"
claude --headless "Run the night shift. Read NIGHT_SHIFT_MANIFEST.yaml, follow night-shift skill RUN mode."
Per task (dependency order):
CHECK compaction gate — stop if context > threshold (default 75%)
READ TaskGet → full context
CLAIM TaskUpdate status=in_progress
PLAN write scratchpad execution plan before touching files
EXEC do the work
TEST verify DONE criterion
DONE TaskUpdate completed|blocked + finalize scratchpad
LOG append to DIGEST
Completion signals: scratchpad must contain TASK_COMPLETE or TASK_BLOCKED: <reason>.
Neither after one retry → auto-blocked. Two iterations >85% identical scratchpad → TASK_BLOCKED: stuck-loop.
Compaction gate: <60% proceed · 60–75% warn · >75% stop, write remaining tasks to DIGEST.
Details: references/compaction-gate.md. Script: scripts/check-context.sh.
DIGEST at .claude/NIGHT_SHIFT_DIGEST.md. Format: references/digest-format.md.
Mode 3: AUDIT LOOP
claude --dangerously-skip-permissions --headless \
"Run night-shift audit loop. Read NIGHT_SHIFT_MANIFEST.yaml, follow night-shift skill AUDIT LOOP mode."
Loop:
AUDIT read spec + codebase → score → write .claude/audit/audit-NNN.md
CHECK score ≥ threshold → COMPLETE
same High finding stuck 2 cycles → BLOCKED
score regressed → git revert HEAD → BLOCKED
TASKS convert findings to risk-tagged tasks
EXEC implement safe/review tasks (code only — hard stops above apply)
TEST run test suite — stop if regression
COMMIT git commit per task "[night-shift] <title> (audit-NNN)"
→ repeat
Auditor
Fresh spawn each iteration. Read-only — writes only to .claude/audit/.
Reads: spec file (spec_path in MANIFEST), AGENTS.md/CLAUDE.md/README.md,
targeted codebase reads, previous audit for delta.
Output .claude/audit/audit-NNN.md:
# Audit NNN — <timestamp>
## Scores
overall: X.X/10
<dimension>: X.X/10 ...
## Findings
[High|Medium|Low] <finding> — file:line
## Task List For Executor
- <task with WHAT / FILES / DONE / risk tag>
## Delta from Previous
improved · regressed · new
## Exit Condition
THRESHOLD_MET | CONTINUE | BLOCKED: <reason>
Rules: cite file+line for every finding · task list must be actionable · always write exit signal.
Executor
Implements auditor's task list. Hard stops above apply unconditionally.
Per task: run tests · commit · if task needs infra to verify → TASK_BLOCKED: requires-infra.
MANIFEST config for audit loop
audit_loop:
spec_path: SPEC.MD
score_threshold: 9.5
max_iterations: 10
dimensions:
- authorization_security
- api_completeness
- domain_model
- testing_coverage
- production_polish
Resuming
export CLAUDE_CODE_TASK_LIST_ID="<id from manifest>"
claude
> Show me .claude/NIGHT_SHIFT_DIGEST.md
> Show me .claude/audit/
Same TASK_LIST_ID across sessions — completed tasks skipped, audit loop resumes from last NNN.
References (load on demand)
references/compaction-gate.md— context % from JSONL transcriptsreferences/digest-format.md— DIGEST.md formatreferences/manifest-schema.yaml— full MANIFEST schemareferences/task-templates.md— task templates by work typescripts/check-context.sh— context budget checkerscripts/launch-night-shift.sh— cron launcher