odw_executor_matrix

A reference describing how Open Dynamic Workflows runs coding agents through command-line executors. An executor is the program that receives a task and returns the agent’s result.

In plain words
What is it for?
Use it to configure or evaluate executors for Claude, Codex, Grok, and other supported coding-agent command-line tools.
Why use it?
It clarifies which coding-agent command-line tools can run workflow tasks and how their output is handled.

Agent

Install

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.

agentmods
npx agentmods add agents/atebites-hub/pjtemplate/odw_executor_matrix
Clone the repo
git clone --depth 1 https://github.com/atebites-hub/PJTemplate
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,885 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.01885
Opus 5 $0.00000 $0.00942
Sonnet 5 $0.00000 $0.00377
Haiku 4.5 $0.00000 $0.00188

Measured 2d ago against content hash c511db0d54dd, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

odw_executor_matrix 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 2d 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.

docs/agents/odw_executor_matrix.md · 86 lines

How it starts

The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.

open-dynamic-workflows — harness / executor matrix

Process doc (not one of the 10 core product documents). Records which coding harnesses this template cares about, whether they expose a headless CLI that can back an ODW Executor, and what is bundled today vs feasible to add. Reviewed against local CLIs + primary docs (2026-07).

How ODW plugs into a harness

ODW does not embed models. Each agent() call names an executor string; the host supplies RunOptions.executors: Record<string, Executor>. An Executor is (ExecOptions) => Promise<ExecResult> — typically spawn a coding-agent CLI, feed the prompt (stdin or argv), parse stdout/JSONL, return final text / structured output.

Bundled in vendor/open-dynamic-workflows (via builtinExecutors):

Name Binary / argv sketch Notes
claude claude --print --output-format=stream-json --verbose --permission-mode acceptEdits Prompt on stdin; stream-json reducer
codex codex exec --json --skip-git-repo-check --color never --sandbox workspace-write - Prompt on stdin (-); JSONL reducer
grok grok --output-format streaming-json --permission-mode acceptEdits --cwd <cwd> --prompt-file <tmp> Grok Build headless; optional --json-schema; never bypassPermissions
cursor cursor-agent --print --output-format stream-json --force --workspace <cwd> <prompt> Command is cursor-agent (not bare agent, which may be Grok’s shim). No native schema flag — schema → prompt instruct + JSON.parse

New harnesses = new adapter under executor/<name>/ using makeSubprocessExecutor (or a pure-function/SDK executor). Do not fork ODW in-tree for adapters unless upstreaming; prefer PR to imsai-sh/open-dynamic-workflows or a thin project wrapper that builds a custom executors map and calls runWorkflow.

Matrix (template-relevant harnesses)

Status legend:

  • Bundled — ships in ODW builtinExecutors
  • Headless-ready — documented / local CLI supports one-shot non-interactive run; adapter not written yet
  • Partial — headless exists but output contract is weak/unstable for reliable resume journals
  • IDE-only / no CLI — not a good ODW leaf executor without another path
  • Unverified — conflicting or thin primary docs

Read the full file on GitHub · 86 lines

Changes

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.

  1. 2d ago First seen · 86 lines · 0 tokens per session scan A c511db0d54dd

Subscribe to this mod's changes

odw_executor_matrix is an agent published in the GitHub repository atebites-hub/PJTemplate (2 stars, last pushed 10d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,885 tokens. 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.