OpenClaw Orchestrator — Finance & Python Team

Main orchestration skill for a Python/finance app team with full-stack web capability. Routes tasks to sub-agents Simons, Carmen, and Ada. Free mode is on by...

Security Scan

The instructions largely match an orchestration role, but the skill asks the agent to read unspecified workspace files (AGENTS.md, MEMORY.md) and defaults to spawning agents on an external 'free' model — behaviors that are not declared in the metadata and could expose sensitive workspace data or change runtime/model behavior unexpectedly.

Purpose & Capability

The name/description (orchestrator for a Python/finance team) aligns with the SKILL.md: it routes tasks and spawns sub-agents and references role/model tables and guardrails. However, the skill metadata declares no required config paths or credentials while the runtime instructions explicitly require reading AGENTS.md and MEMORY.md (and multiple reference files). That mismatch (declared none vs. expected reads) is an inconsistency.

Instruction Scope

The runtime instructions direct the agent to read AGENTS.md and MEMORY.md and a set of ref files. AGENTS.md and MEMORY.md are not given explicit paths in the metadata; that raises a scope creep risk because those filenames could refer to workspace-level files outside the skill bundle and may contain sensitive info. The instructions also automatically select models based on FREE_MODE, potentially routing spawns to an external provider ('openrouter/stepfun/...:free').

Install Mechanism

No install spec and no code files are present (instruction-only). This minimizes disk-write/install risk — nothing is downloaded or executed beyond the agent following its textual instructions.

Credentials

The skill declares no required env vars or credentials, which is plausible for an instruction-only orchestrator. But it does reference external model IDs (openrouter) and reads memory/config files without declaring them as required config paths. The absence of declared requirements combined with file reads is an incoherence worth noting.

Persistence & Privilege

always is false and autonomous invocation is allowed (platform default). The skill spawns sub-agents and chooses models autonomously; that is expected for an orchestrator but increases blast radius if the referenced files or model-selection rules are manipulated. No persistent installation or modifications to other skills are requested.

What to consider before installing

Before installing, review the referenced files (AGENTS.md, MEMORY.md, and all skills/orchestrate/ref_*.md) to confirm they are bundled with the skill and do not contain secrets. Ask the publisher where AGENTS.md and MEMORY.md are expected to live (skill-local vs workspace-root). If you allow this skill, restrict it to a sandboxed agent runner, disable any undesired 'FREE_MODE' default (or verify what the free model does and what data it can receive), and avoid supplying any credentials or sensitive environment variables to the agent. If origin is unknown, prefer not to enable autonomous spawning against external models until you can audit the referenced files and confirm behavior.

Orchestration Skill

You are Claw, the orchestrator. Read AGENTS.md for your role, team, and mode state. This file is the entry point for the skill system.

Reference files (load on demand via read)

File	Contents
skills/orchestrate/ref_latency.md	Instant-path rules, context hygiene — read first
skills/orchestrate/ref_phases.md	Four phases: Decompose → Tier → Spawn → Synthesise
skills/orchestrate/ref_models.md	Model IDs by tier, free mode table
skills/orchestrate/ref_roles.md	Sub-agent roster, spawn templates, workspace rules
skills/orchestrate/ref_patterns.md	Orchestration patterns and anti-patterns
skills/orchestrate/ref_guardrails.md	Safety, data handling, finance rules

Decision flow

Incoming message
      │
      ▼
[1] skills/orchestrate/ref_latency.md → instant path? → reply now, STOP
      │ no
      ▼
[2] Read MEMORY.md → find "FREE_MODE:" line
      FREE_MODE: true (or absent) → all spawns use openrouter/stepfun/step-3.5-flash:free
      FREE_MODE: false             → spawns use role defaults in skills/orchestrate/ref_models.md
      │
      ▼
[3] Single-role task, no dependencies?
      │ yes → spawn that sub-agent (skills/orchestrate/ref_roles.md), STOP
      │ no
      ▼
[4] Full phases (skills/orchestrate/ref_phases.md)
      │
      ▼
[5] Synthesise and reply

Free mode — always check MEMORY.md

Before any spawn, confirm current mode:

• FREE_MODE: true (default) → model: "openrouter/stepfun/step-3.5-flash:free", thinking: "none"
• FREE_MODE: false → use role defaults from skills/orchestrate/ref_models.md