Honey Loop
A loop is the one place token waste compounds: cost = per_tick × ticks. The
honey levers still apply inside each tick (less code, less prose), but a loop adds
two leaks the single-shot levers don't cover — re-paying for context every
wake-up, and re-doing work that didn't change. Cut those.
The five levers
Pace to the cache, skip the dead zone. The prompt cache TTL is 5 min.
< 270s → context stays warm; right for actively polling external state.≥ 1200s → one cache miss amortized over a long idle wait; the default for
idle ticks (use 1200–1800s).~300s is the worst choice — pays the cache miss without amortizing it.
Never round to "5 minutes." Drop to 270s or commit to 1200s+.
Event-driven over polling. Harness-tracked work (background Bash, Agent,
Workflow) re-invokes you automatically on completion. Don't schedule a short
tick to poll it — that re-reads full context for nothing. Set a long fallback
heartbeat (1200s+) and let the completion notification drive you. Poll only
external state the harness can't see (CI, a deploy, a remote queue), at a
cadence matched to how fast that state actually changes.
No-change short-circuit. Re-generating output is the bulk of per-tick cost.
Check cheaply first (a hash, a timestamp, a git rev-parse, a status field). If
nothing changed: emit one line and reschedule. Don't redo the task, don't
re-read the world, don't write a report.
Carry a compact state handle, don't re-derive. Keep what the loop has done
so far in a small file (.honey-loop/state.json or a <<honey:HASH>> stash via
eso stash), and read that each tick — not the full history re-discovered from
scratch. Lever 3b: the cheapest input token is the one that never re-enters context.
Define done; then stop. An unbounded loop is unbounded cost. State the exit
condition up front (target reached, N empty ticks, deadline). When met, end the
loop — omit the wake-up reschedule. "Loop forever" is a bug, not a feature.
Per-tick output: ultra by default
A loop tick is an agent-to-self / agent-to-log message, not a deliverable. Default
to ultra: status line only, near-zero prose. One line is enough —
tick 4/∞ · no change · next 1800s. Step up to a full report only on the tick
where something actually needs the user (state changed, a decision, an error).
Don't cut
- The stop/exit check — dropping it is how a loop runs all night.
- Error surfacing — a tick that hit an error reports it; silent failure in a
loop hides for hours.
- The cheap pre-check in lever 3 — skipping it to "just redo it" defeats the point.
1---2name: honey-loop3description: Cost discipline for recurring /loop runs. A loop multiplies per-iteration cost by iteration count, so waste compounds. Encodes cache-aware pacing (avoid the 300s dead zone), event-driven-over-polling, a no-change short-circuit, a compact state handle carried between ticks, and a stop condition. Use when a /loop is running or being set up - especially long-lived or frequently-firing ones.4---56# Honey Loop78A loop is the one place token waste *compounds*: `cost = per_tick × ticks`. The9honey levers still apply inside each tick (less code, less prose), but a loop adds10two leaks the single-shot levers don't cover — **re-paying for context** every11wake-up, and **re-doing work** that didn't change. Cut those.1213## The five levers14151. **Pace to the cache, skip the dead zone.** The prompt cache TTL is 5 min.16 - `< 270s` → context stays warm; right for actively polling external state.17 - `≥ 1200s` → one cache miss amortized over a long idle wait; the default for18 idle ticks (use **1200–1800s**).19 - **`~300s` is the worst choice** — pays the cache miss without amortizing it.20 Never round to "5 minutes." Drop to 270s or commit to 1200s+.21222. **Event-driven over polling.** Harness-tracked work (background `Bash`, `Agent`,23 `Workflow`) re-invokes you automatically on completion. Don't schedule a short24 tick to poll it — that re-reads full context for nothing. Set a long fallback25 heartbeat (`1200s+`) and let the completion notification drive you. Poll only26 external state the harness can't see (CI, a deploy, a remote queue), at a27 cadence matched to how fast that state actually changes.28293. **No-change short-circuit.** Re-generating output is the bulk of per-tick cost.30 Check cheaply first (a hash, a timestamp, a `git rev-parse`, a status field). If31 nothing changed: emit **one line** and reschedule. Don't redo the task, don't32 re-read the world, don't write a report.33344. **Carry a compact state handle, don't re-derive.** Keep what the loop has done35 so far in a small file (`.honey-loop/state.json` or a `<<honey:HASH>>` stash via36 `eso stash`), and read *that* each tick — not the full history re-discovered from37 scratch. Lever 3b: the cheapest input token is the one that never re-enters context.38395. **Define done; then stop.** An unbounded loop is unbounded cost. State the exit40 condition up front (target reached, N empty ticks, deadline). When met, end the41 loop — omit the wake-up reschedule. "Loop forever" is a bug, not a feature.4243## Per-tick output: ultra by default4445A loop tick is an agent-to-self / agent-to-log message, not a deliverable. Default46to **ultra**: status line only, near-zero prose. One line is enough —47`tick 4/∞ · no change · next 1800s`. Step up to a full report **only** on the tick48where something actually needs the user (state changed, a decision, an error).4950## Don't cut5152- The **stop/exit check** — dropping it is how a loop runs all night.53- **Error surfacing** — a tick that hit an error reports it; silent failure in a54 loop hides for hours.55- The **cheap pre-check** in lever 3 — skipping it to "just redo it" defeats the point.
S1lverElixir