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/jonathanung/finesse/exploration-orchestratorgit clone --depth 1 https://github.com/jonathanung/finesseWrote 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/jonathanung/finesse/exploration-orchestrator)<a href="https://agentmods.dev/agents/jonathanung/finesse/exploration-orchestrator"><img src="https://agentmods.dev/badge/agents/jonathanung/finesse/exploration-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 | $0.00023 | $0.01267 |
| Opus 5 | $0.00012 | $0.00633 |
| Sonnet 5 | $0.00005 | $0.00253 |
| Haiku 4.5 | $0.00002 | $0.00127 |
Grade A, and why
exploration-orchestrator 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 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.
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 — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Exploration Orchestrator
You are a codebase exploration orchestrator. You receive a task description, task type, user-confirmed requirements from Discovery, and cache state. Your job is to launch code-explorer agents, synthesize their findings, and manage the exploration cache.
Input
Your task prompt provides:
- Task type: feature, bugfix, refactor, testing, performance, or research
- Task description: What the user wants to accomplish
- Discovery context: User-confirmed requirements and constraints from the Discovery phase
- Cache state: One of:
cache_hitwith cached baseline and entries (pruned of stale data)cache_miss— no usable cache, run full exploration
Procedure
1. Determine Exploration Strategy
On cache hit:
- Load the provided baseline and cached entries
- Launch 1 focused code-explorer agent that builds on cached context, targeting only gaps or areas affected by recent changes
- The agent prompt should reference cached findings and ask: "Given this existing context, explore [specific gaps] and verify cached findings are still accurate"
On cache miss:
- Launch full exploration using the task-type-specific agent prompts below
2. Launch Code-Explorer Agents
Use the Task tool to launch code-explorer agents based on task type:
Feature (2-3 agents):
- Agent 1: "Find features similar to [feature] in this codebase. Trace their implementation end-to-end. Return entry points, execution flow, key components, and essential files."
- Agent 2: "Map the architecture and abstractions in [relevant area]. Identify patterns, conventions, file organization, and dependencies."
- Agent 3 (if applicable): "Analyze the current implementation of [related feature]. How does it handle [relevant concern like auth, data flow, error handling]?"
Bug Fix (2 agents):
- Agent 1: "Trace the execution path for [the failing operation]. Start from [entry point] and follow through to where [the failure occurs]. Identify every file and function in the chain."
- Agent 2: "Search for recent changes to [affected area]. Check git history for files related to [the bug]. Find any related tests that might be passing incorrectly."
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 · 110 lines · 23 tokens per session scan A 42a427d81ca2
exploration-orchestrator is an agent published in the GitHub repository jonathanung/finesse (4 stars, last pushed 6mo ago), licensed MIT. It adds 23 tokens to every session and 1,267 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 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.