agentmods.dev

deepcrew-ai: Instructions file for Claude Code

CLAUDE.md

deepcrew-ai CLAUDE.md is an instructions file for Claude Code from Aayush-Joshi-01/deepcrew-ai. It costs 1,931 tokens per session, scanned A, original, MIT.

Repository guidance for deepcrew-ai, an asynchronous Python library that runs AI agents through models such as OpenAI, Anthropic, Gemini, or local Ollama models.

In plain words
What is it for?
Use it when working on agents, tool execution, parallel coordination, refinement and verification loops, streaming, memory, or multi-agent result synthesis.
Why use it?
It maps the library’s main components and conventions so coding agents can change the right part of the system safely.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions subagents.

This is Aayush-Joshi-01/deepcrew-ai's own configuration. It tells Claude Code how to work on deepcrew-ai 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 deepcrew-ai configures →

not rated 3repo changed 12d ago A scan 1,931 tokens original MIT
View source ↗ Aayush-Joshi-01/deepcrew-ai deepcrew-ai.aayushjoshi.dev
Reuse

Borrowing it

Nothing to install: this file belongs to Aayush-Joshi-01/deepcrew-ai. 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/Aayush-Joshi-01/deepcrew-ai/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/Aayush-Joshi-01/deepcrew-ai

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 deepcrew-ai CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/aayush-joshi-01/deepcrew-ai/claude-md/github.svg)](https://agentmods.dev/instructions/aayush-joshi-01/deepcrew-ai/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/aayush-joshi-01/deepcrew-ai/claude-md"><img src="https://agentmods.dev/badge/instructions/aayush-joshi-01/deepcrew-ai/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 deepcrew-ai CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/aayush-joshi-01/deepcrew-ai/claude-md"><img src="https://agentmods.dev/badge/instructions/aayush-joshi-01/deepcrew-ai/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 1,931 This file is loaded in full into every session.
When invoked 1,931 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 $0.01931 $0.01931
Opus 5.5 $0.00772 $0.00772
Sonnet 5.5 $0.00386 $0.00386
Haiku 4.5 $0.00193 $0.00193

Measured 12d ago against content hash bf24878d094c, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

deepcrew-ai 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 12d 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 · 114 lines

How it starts

The opening of the file, as written. The whole thing — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

Guidance for coding agents (Claude Code and others) working in this repository.

What this is

deepcrew-ai (import name deepcrew) is an async, LiteLLM-backed multi-agent library. Its distinguishing features are a self-improving refinement loop (verifier-scored, adaptive, self-consistency branching, skill distillation), bounded recursive agent spawning, APEX (Assembly & Precision Executor) multi-agent synthesis, and true token streaming with selectable visibility (StreamPolicy).

Architecture map

  • src/deepcrew/agent.py — Agent dataclass: model, system prompt, tools, memory, hooks, response_model, retry/fallback config.
  • src/deepcrew/runner.py — run_agent(): the core agentic loop (stream → buffer tool calls → execute in parallel → repeat). Delegates to loop.py when Agent.loop_config is set.
  • src/deepcrew/orchestrator.py — Orchestrator: NIA (Navigator & Intelligence Allocator), the router LLM, picks single-agent or parallel fan-out, then apex.py's APEXSynthesizer (APEX, Assembly & Precision Executor) merges parallel results.
  • src/deepcrew/loop.py — run_agent_loop(): outer refinement loop driven by LoopConfig — verifier scoring (verifier.py), adaptive early-stop, self-consistency branching, and Voyager-style skill distillation into skills/registry.py.
  • src/deepcrew/spawner.py — bounded recursive sub-agent spawning (spawn_agent meta-tool), hard-capped by max_spawn_depth.
  • src/deepcrew/procedural_memory.py — durable, evolving playbook layered on any MemoryProvider. Optional reflect_model swaps the heuristic curate() for one LLM call.
  • src/deepcrew/episodic_memory.py — EpisodicMemory / Episode: a ring-buffer log of past runs layered on any MemoryProvider, with dependency-free similarity+recency recall(). Agent.episodic_memory drives an automatic recall-before / record-after cycle in runner.py (_run_agent_turns recalls; the public run_agent and loop.py record — one episode per loop iteration via record_agent_episode).
  • src/deepcrew/memory/ — MemoryProvider ABC + InMemoryProvider, FileMemoryProvider, RedisMemoryProvider (lazy-imported; needs the redis extra).
  • src/deepcrew/content.py — multimodal input: image(), pdf(), user_message() build OpenAI content blocks; extract_text() is the canonical way to pull plain text out of a message whose content may be a string or a block list.
  • src/deepcrew/stream.py — StreamEvent/EventType, StreamPolicy (chat/standard/verbose presets), filter_stream(). Filtering is view-only — it never changes what actually executes.
  • src/deepcrew/logging_config.py — general-purpose logging layer: LoggingConfig, configure_logging() (named profiles dev/prod/verbose/quiet, env-var toggles DEEPCREW_LOG*), and get_logger() (a StructuredAdapter — kwargs become structured fields; public API for user components). Colour via the optional rich ([log] extra), plain ANSI fallback otherwise. Off by default (NullHandler only). Independent of OTel and of the StreamEvent queue.
  • src/deepcrew/hooks.py — AgentHooks: human-in-the-loop interception (approve_tool can deny a tool call outright). Distinct from the observe-only event stream.
  • src/deepcrew/workflow.py — WorkflowBuilder: explicit DAG of agents, Kahn's-algorithm level scheduling so independent nodes run in parallel.
  • src/deepcrew/mcp/ — MCP client transports (stdio, SSE, streamable HTTP) + MCPManager.
  • src/deepcrew/integrations/fastapi.py — optional (fastapi extra) drop-in SSE router.
  • src/deepcrew/cli/ — deepcrew run <workflow.yaml> / deepcrew agents list.

Read the full file on GitHub · 114 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. 12d ago Changed · +26 lines · +552 tokens per session bf24878d094c
  2. 2mo ago First seen · 88 lines · 1,379 tokens per session scan A a717a9dcc76e

Subscribe to this mod's changes

deepcrew-ai CLAUDE.md is an instructions file published in the GitHub repository Aayush-Joshi-01/deepcrew-ai (3 stars, last pushed 13d ago), licensed MIT. It adds 1,931 tokens to every session, about $0.0077 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 2026-08-31.

Autonomous AgentsBuilding agentsagent-frameworkagentic-aiai-agentsanthropicasyncio
Related

deepcrew-ai CLAUDE.md compared with instructions from other repositories

deepcrew-ai CLAUDE.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
deepcrew-ai CLAUDE.mdAayush-Joshi-01/deepcrew-aiInstructions file313d agoMITA1,931Claude Code
agentforge-py copilot-instructions.mdScaffoldic/agentforge-pyInstructions file101mo agoApache-2.0A101GitHub Copilot
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA9,719GitHub Copilot
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA5,516GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143kyesterdayMITA7,461Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchainInstructions file149ktodayMITA4,680Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA3,093GitHub Copilot