Parallel Ops — router
You have parallel or recurring agent work and don't know which skill owns it.
Six skills orbit this space and their names alone don't disambiguate. This
router owns cross-family discovery; read the table, jump to the one skill
you need, and stop reading here.
The decision table
| You want |
Go to |
Not this, because |
| Parallel subtasks run by your session's OWN provider and model tier, in-process |
native Workflow tool / Agent subagents (isolation: worktree) |
not a fleet skill at all — no dedicated skill needed |
| Work done by a CHEAPER brain than your session (GLM/Haiku/Sonnet-under-Opus) — one subtask or a whole fan-out, all one brain type |
fleet-worker |
fleetflow is overkill when every worker runs the same brain |
| DIFFERENT brains per work class in one run, or cross-provider dissent in verify (e.g. Codex or Grok refutes GLM) |
fleetflow (own repo — 0xDarkMatter/fleetflow, mounted at ~/.claude/skills/fleetflow) |
fleet-worker runs one brain type per run (any provider, but not mixed). fleetflow is also the home for a Grok (xAI) worker — its own binary, not a claude -p brain |
| Work that RECURS on a schedule across sessions — cron, routine, unattended ticks |
loop-ops |
iterate is one continuous session, not a schedule |
| Drive ONE mechanical metric to a target in one continuous session (even a long overnight one) |
iterate |
loop-ops is the scheduler around sessions, not the session itself |
| Land/merge branches that parallel work produced |
fleet-ops |
the terminus for every branch-producing row above (in-process subagents and prompt authoring produce no branches) |
| Author a static expert-agent prompt FILE (not a runtime worker) |
spawn |
listed only to catch the name collision with "spawn workers" |
Tie-breakers for the two classic overlaps: "many files, cheap models" is
brain economics, not count — cheaper brain → fleet-worker, own brain →
native. "Run overnight until X" is session shape, not duration — one
continuous run → iterate; scheduled re-entry across sessions → loop-ops.
Two axes that confuse cold agents
Spawn vs. land. fleet-worker and fleetflow spawn workers and produce
branches; fleet-ops lands those branches through a test-gated queue.
Every branch-producing lane ends at fleet-ops regardless of how it was
spawned — agent team, background agent, claude -p worker, or human.
Inner loop vs. outer loop. iterate is the inner loop: one session,
one metric, git as memory, runs until a stop condition. loop-ops is the
outer loop: the scheduler and risk-tier discipline that decides when
and whether to fire a run (inner loop or otherwise) unattended.
Composition chain
iterate (inner loop) → loop-ops (outer loop / scheduler)
↓
fleet-worker / fleetflow (spawn workers)
↓
fleet-ops (land branches)
Not every task uses the whole chain — most use exactly one link. Read the
table above first; only compose when the task genuinely spans spawn +
land or inner + outer.
See also
- fleet-ops — landing discipline: test-gated queue, pre-land scrub, auto-rebase, revert
- fleet-worker — one cheap headless worker (GLM, Sonnet, Haiku) fanned out and gated
- fleetflow — heterogeneous cross-provider fleet (GLM + Codex + Grok + Pi + Anthropic); extracted to its own repo at 0xDarkMatter/fleetflow, still invocable as
/fleetflow via the junction at ~/.claude/skills/fleetflow
- loop-ops — outer-loop design: risk tiers, kill switch, scheduling
- iterate — autonomous single-metric improvement loop
- spawn — generates expert-agent prompt files (authoring, not runtime)
1---2name: parallel-ops3description: Router for parallel or recurring agent work across six skills. Covers: parallel agents, fan out work, delegate to workers, run overnight, scheduled loop, land branches, mixed-model fleet, orchestrate workers, background agents at scale. Triggers on: which skill for parallel work, fan out agents, spawn workers, run this overnight, schedule a loop, land my branches, heterogeneous fleet, delegate to cheaper model, autonomous loop.4license: MIT5---67# Parallel Ops — router89You have parallel or recurring agent work and don't know which skill owns it.10Six skills orbit this space and their names alone don't disambiguate. This11router owns cross-family discovery; read the table, jump to the one skill12you need, and stop reading here.1314## The decision table1516| You want | Go to | Not this, because |17|---|---|---|18| Parallel subtasks run by your session's OWN provider and model tier, in-process | native Workflow tool / `Agent` subagents (`isolation: worktree`) | not a fleet skill at all — no dedicated skill needed |19| Work done by a CHEAPER brain than your session (GLM/Haiku/Sonnet-under-Opus) — one subtask or a whole fan-out, all one brain type | [fleet-worker](../fleet-worker/) | fleetflow is overkill when every worker runs the same brain |20| DIFFERENT brains per work class in one run, or cross-provider dissent in verify (e.g. Codex or Grok refutes GLM) | fleetflow (own repo — [0xDarkMatter/fleetflow](https://github.com/0xDarkMatter/fleetflow), mounted at `~/.claude/skills/fleetflow`) | [fleet-worker](../fleet-worker/) runs one brain type per run (any provider, but not mixed). fleetflow is also the home for a **Grok** (xAI) worker — its own binary, not a `claude -p` brain |21| Work that RECURS on a schedule across sessions — cron, routine, unattended ticks | [loop-ops](../loop-ops/) | [iterate](../iterate/) is one continuous session, not a schedule |22| Drive ONE mechanical metric to a target in one continuous session (even a long overnight one) | [iterate](../iterate/) | [loop-ops](../loop-ops/) is the scheduler *around* sessions, not the session itself |23| Land/merge branches that parallel work produced | [fleet-ops](../fleet-ops/) | the terminus for every branch-producing row above (in-process subagents and prompt authoring produce no branches) |24| Author a static expert-agent prompt FILE (not a runtime worker) | [spawn](../spawn/) | listed only to catch the name collision with "spawn workers" |2526Tie-breakers for the two classic overlaps: "many files, cheap models" is27**brain economics, not count** — cheaper brain → fleet-worker, own brain →28native. "Run overnight until X" is **session shape, not duration** — one29continuous run → iterate; scheduled re-entry across sessions → loop-ops.3031## Two axes that confuse cold agents3233**Spawn vs. land.** fleet-worker and fleetflow *spawn* workers and produce34branches; fleet-ops *lands* those branches through a test-gated queue.35Every branch-producing lane ends at fleet-ops regardless of how it was36spawned — agent team, background agent, `claude -p` worker, or human.3738**Inner loop vs. outer loop.** iterate is the *inner* loop: one session,39one metric, git as memory, runs until a stop condition. loop-ops is the40*outer* loop: the scheduler and risk-tier discipline that decides when41and whether to fire a run (inner loop or otherwise) unattended.4243## Composition chain4445```46iterate (inner loop) → loop-ops (outer loop / scheduler)47 ↓48 fleet-worker / fleetflow (spawn workers)49 ↓50 fleet-ops (land branches)51```5253Not every task uses the whole chain — most use exactly one link. Read the54table above first; only compose when the task genuinely spans spawn +55land or inner + outer.5657## See also5859- [fleet-ops](../fleet-ops/) — landing discipline: test-gated queue, pre-land scrub, auto-rebase, revert60- [fleet-worker](../fleet-worker/) — one cheap headless worker (GLM, Sonnet, Haiku) fanned out and gated61- fleetflow — heterogeneous cross-provider fleet (GLM + Codex + Grok + Pi + Anthropic); extracted to its own repo at [0xDarkMatter/fleetflow](https://github.com/0xDarkMatter/fleetflow), still invocable as `/fleetflow` via the junction at `~/.claude/skills/fleetflow`62- [loop-ops](../loop-ops/) — outer-loop design: risk tiers, kill switch, scheduling63- [iterate](../iterate/) — autonomous single-metric improvement loop64- [spawn](../spawn/) — generates expert-agent prompt files (authoring, not runtime)
0xDarkMatter