Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/wazimmerman/zimster/dispatching-parallel-agentsnpx skills add wazimmerman/zimster --skill dispatching-parallel-agentsgit clone --depth 1 https://github.com/wazimmerman/zimsterWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/wazimmerman/zimster/dispatching-parallel-agents)<a href="https://agentmods.dev/skills/wazimmerman/zimster/dispatching-parallel-agents"><img src="https://agentmods.dev/badge/skills/wazimmerman/zimster/dispatching-parallel-agents.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00023 | $0.00595 |
| Opus 5 | $0.00012 | $0.00298 |
| Sonnet 5 | $0.00005 | $0.00119 |
| Haiku 4.5 | $0.00002 | $0.00060 |
Grade A, and why
dispatching-parallel-agents scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dispatching Parallel Agents
Parallelism is useful when workstreams do not depend on the same evolving state. It is not a default response to a long plan.
Independence test
Parallel work is safe only when all are true:
- agents can complete without decisions produced by the other;
- file or component ownership does not overlap;
- each output has a clear integration contract;
- failures can be evaluated independently;
- concurrent repository operations will not corrupt branch or index state.
If two tasks share a lifecycle, state machine, public interface, migration, or frequently edited files, keep one persistent owner.
Default limit
Use a maximum of two parallel implementation agents. Read-only investigations may be batched when the harness supports it, but each still needs a bounded question and output contract.
Subagents must not spawn subagents. The root owner remains responsible for integration, verification, and completion claims.
Dispatch contract
First record whether delegation is selected and why it materially improves the task. Also record why inline execution is less appropriate. An inexpensive or available model is never evidence that delegation is useful. If delegation is not selected, stop: no model proposal or dispatch is permitted.
Every selected delegation states:
- one outcome or question;
- owned files or read-only boundary;
- binding interfaces and constraints;
- dependency cone, tool restrictions, stop condition, and owner acceptance proof;
- turn or time budget;
- exact return format;
- where detailed artifacts should be written;
- prohibition on committing outside the assigned scope.
Only then create a plan-time advisory or dispatch-time authoritative proposal
using the harness-neutral economy, balanced, expert, or inherit class.
Resolve optional user mappings and current capability/catalog evidence at
dispatch time. Record requested and effective model/effort separately. A
strict-cost request that cannot prove enforcement returns optional work to the
owner or blocks required review; it never silently inherits.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 83 lines · 23 tokens per session scan A 1dd97ac631cb
dispatching-parallel-agents is a skill published in the GitHub repository wazimmerman/zimster (2 stars, last pushed 3d ago), licensed MIT. It adds 23 tokens to every session and 595 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
build-test
Run the project's build / typecheck / lint / test commands and emit the build.passing + tests.passing signals devloop convergence reads.
test-driven-development
Drive changes with Red→Green→Refactor; ensure behavior is verifiable and regression-safe.
rudder
Use locally captured coding-session intent to resolve a device-local behavioral spec, generate focused tests from that spec, implement the smallest production changes through red-green TDD, and verify coverage with the repository's native tooling. Use when the user asks to run Rudder, create or regenerate tests for…
TDD 开发纪律助手
引导开发过程遵循测试先行、红绿重构和回归验证纪律。.
treework-manual
Use TreeWork Manual when the user asks for it, or when the Agent judges that organizing ongoing, multi-part work as a persistent Markdown Tree would improve direction, continuity, recovery, or handoff. Adapt the Tree's depth and document detail to the user's needs, domain, and situation rather than a fixed threshold.
treework
Use TreeWork, a tree-guided development plugin for complex or evolving software projects that need staged alignment, pre-coding technical Specs, declarative project-tree planning, branch-scoped Git worktrees, durable recovery, and protected completion. It teaches the Agent to move through development along the Tree…