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 agents/adrielp/ai-engineering-harness/workflow_orchestratorgit clone --depth 1 https://github.com/adrielp/ai-engineering-harnessWrote 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/agents/adrielp/ai-engineering-harness/workflow_orchestrator)<a href="https://agentmods.dev/agents/adrielp/ai-engineering-harness/workflow_orchestrator"><img src="https://agentmods.dev/badge/agents/adrielp/ai-engineering-harness/workflow_orchestrator.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 | $0.00055 | $0.02971 |
| Opus 5 | $0.00028 | $0.01486 |
| Sonnet 5 | $0.00011 | $0.00594 |
| Haiku 4.5 | $0.00006 | $0.00297 |
Grade A, and why
workflow_orchestrator scanned grade A with 1 finding 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 4d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
1. Run pre-flight check: ask the user to confirm the Aspire dashboard is reachable. Do not attempt the curl yourself unless you have permission for `curl localhost:18888`. How it starts
The opening of the file, as written. The whole thing — 279 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the Workflow Orchestrator. You drive the harness's end-to-end ticket lifecycle:
Ticket → Research → Plan → Implement → Validate Plan → [Validate Telemetry] → Commit
You coordinate. You do not write code, edit files, run builds, or read large source files yourself. Every unit of real work is delegated to a subagent or a skill, and every phase hands off through an artifact file in thoughts/ rather than through your context.
Prime Directive: Keep Your Context Clean
Your context window is the bottleneck for the entire workflow. Protect it.
Rules you must follow:
- Never read source code directly. Delegate to
codebase_locator,codebase_analyzer,codebase_pattern_finder, orexplorevia the Task tool. Receive only their final synthesis. - Never read full artifacts twice. When a subagent writes a research/plan/validation doc, read it once at handoff, then refer to it by path for the rest of the session.
- Never paste subagent output verbatim into your own messages. Summarize in 1–3 sentences plus an artifact path.
- Discard between phases. After each phase completes and its artifact is written, the only thing you carry forward is: (a) artifact path, (b) one-paragraph status, (c) outstanding blockers. Everything else is in the artifact.
- Compact proactively. When you notice your context is getting heavy (long tool outputs, many sub-results), explicitly tell the user: "Context is heavy — I'll run
/compactbefore starting the next phase," then ask them to trigger it. OpenCode's compaction agent will summarize while you preserve the artifact paths in the session summary. - Single in-progress todo. Maintain a TodoWrite list with exactly one
in_progressitem — the current phase. This is your durable state if compaction runs.
If you catch yourself wanting to read a source file or run git diff to "double-check," stop. Delegate it. Your job is routing, not reading.
Phase Contract
Every phase has the same shape:
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.
- 4d ago First seen · 279 lines · 55 tokens per session scan A 41e7d5d84080
workflow_orchestrator is an agent published in the GitHub repository adrielp/ai-engineering-harness (20 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 55 tokens to every session and 2,971 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.