task-orchestrator: Instructions file for Claude Code

CLAUDE.md

task-orchestrator CLAUDE.md is an instructions file for Claude Code from jpicklyk/task-orchestrator. It costs 6,048 tokens per session, scanned A, original, MIT.

Project instructions for an MCP task orchestrator, a Kotlin server that manages hierarchical work items, dependencies, notes, and role-based workflows. It uses SQLite for stored data and can run from a JAR file or Docker.

In plain words
What is it for?
Working on the Kotlin source, database migrations, task and dependency models, Gradle builds, targeted tests, and Docker-based execution.
Why use it?
They give the agent the project's architecture and the commands needed to build, test, and run it.

Instructions file for Claude Code

Written for Claude Code: SessionStart hook event. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions Claude Code.

This is jpicklyk/task-orchestrator's own configuration. It tells Claude Code how to work on task-orchestrator itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything task-orchestrator configures →

Reuse

Borrowing it

Nothing to install: this file belongs to jpicklyk/task-orchestrator. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/jpicklyk/task-orchestrator/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/jpicklyk/task-orchestrator

Made for: Claude Code.

Wrote 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.

agentmods badge for task-orchestrator CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/jpicklyk/task-orchestrator/claude-md/github.svg)](https://agentmods.dev/instructions/jpicklyk/task-orchestrator/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/jpicklyk/task-orchestrator/claude-md"><img src="https://agentmods.dev/badge/instructions/jpicklyk/task-orchestrator/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for task-orchestrator CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/jpicklyk/task-orchestrator/claude-md"><img src="https://agentmods.dev/badge/instructions/jpicklyk/task-orchestrator/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 6,048 This file is loaded in full into every session.
When invoked 6,048 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.06048 $0.06048
Opus 5.5 $0.02419 $0.02419
Sonnet 5.5 $0.01210 $0.01210
Haiku 4.5 $0.00605 $0.00605

Measured 5d ago against content hash 0a2bba871055, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

task-orchestrator CLAUDE.md 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.

CLAUDE.md · 198 lines

How it starts

The opening of the file, as written. The whole thing — 198 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Project: MCP Task Orchestrator

A Kotlin-based MCP server providing hierarchical work item management with dependency tracking, note schemas, and role-based workflow automation.

Build Commands

# Docker (most common)
docker build -t task-orchestrator:dev .
docker run --rm -i \
  -v mcp-task-data:/app/data \
  -v "$(pwd)"/deploy/global-config/.taskorchestrator:/project/.taskorchestrator:ro \
  -e AGENT_CONFIG_DIR=/project \
  task-orchestrator:dev

Architecture

Source lives under current/.

Package root: io.github.jpicklyk.mcptask.current Source root: current/src/main/kotlin/io/github/jpicklyk/mcptask/current/

Entry point: current/src/main/kotlin/io/github/jpicklyk/mcptask/current/CurrentMain.kt

Composition root: bootstrap/ wires the server. No layer may import it; only CurrentMain.kt and bootstrap itself may (rule R5).

Layer rules (enforced by LayeringTest): domain -> application -> infrastructure -> interfaces (an outer layer may depend on an inner one; the reverse is forbidden) — domain must not import from application/infrastructure/interfaces; application must not import from infrastructure/interfaces; infrastructure must not import from interfaces. infrastructure -> application IS allowed (infrastructure adapters implement application-layer ports); interfaces -> anything is unrestricted; the root package (CurrentMain.kt) is unscoped. Enforced by current/src/test/kotlin/.../architecture/LayeringTest.kt (Konsist), production source only, against the two-way-ratcheted baseline current/src/test/resources/architecture/layering-baseline.txt, which is empty: any violation fails the test, and so would a stale baseline line.

Application sub-packages (enforced acyclic by PackageCycleTest): application/ splits into api (tool request types and handlers' shared context), service (command services, advance), read (read views), run (runs and seats), knowledge (links, related items, documents, search), config (schema/trait resolution, status labels), support (small shared helpers, e.g. IdRef, ScopeGuard, eventRootOf, Cancellation, UnitResults), telemetry, port (storage ports and paging: PageRequest, ListPage, CursorCodec), and the top-level decision, report and upgrade. Dependencies point down this order: api -> service, read, run, knowledge, config, support, telemetry, port (decision and report depend on api, not the reverse); service -> config, knowledge, port, support, telemetry; read -> port, service; run -> config, port, service, support; knowledge -> port, support, telemetry; support -> port; telemetry -> port; config and port depend on no other sub-package. PackageCycleTest fails on any edge that closes a cycle between sub-packages, against current/src/test/resources/architecture/package-cycle-baseline.txt, which is empty. To share a type, move it down to a package every consumer already sits above. Guard tests (ToolSurfaceGuardsTest, ToolSchemaGoldenTest, ApiReferenceFreshnessTest) derive their tool list from buildToolSpecs() (interfaces/shared/tools/ToolRegistry.kt) rather than a hard-coded list, so they automatically cover every tool the production server registers.

Read the full file on GitHub · 198 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. 5d ago Changed · +34 lines · +2,310 tokens per session 0a2bba871055
  2. 20d ago Changed · +1 lines · +54 tokens per session a7d76c7045a9
  3. 1mo ago Changed · +6 lines · +508 tokens per session c902b647c136
  4. 1mo ago Changed · +1 lines · +53 tokens per session 5551658e2b1f
  5. 1mo ago Changed · +18 lines · +576 tokens per session 0df1074dab14
  6. 1mo ago Changed · +2 lines · +178 tokens per session c1e62c916373
  7. 1mo ago Changed · -177 lines · -2,660 tokens per session scan C → A b829e58f3733
  8. 1mo ago Changed · +10 lines · +307 tokens per session 010f6247e8ce
  9. 1mo ago First seen · 303 lines · 4,722 tokens per session scan C 28dfc2d5f1bf

Subscribe to this mod's changes

task-orchestrator CLAUDE.md is an instructions file published in the GitHub repository jpicklyk/task-orchestrator (210 stars, last pushed today), licensed MIT. It adds 6,048 tokens to every session, about $0.0242 per session on Opus 5.5. 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 2026-08-30.

Related

Other instructions, from other repositories