mantis-configure

A setup and validation tool for a Mantis security-review pipeline, including its isolated execution environment, AI model, API connection, and credentials.

In plain words
What is it for?
Use it to create or update workflow.json, select a sandbox and model, set a custom model endpoint, choose reasoning effort, and run preflight checks.
Why use it?
It helps catch configuration or access problems before a security review starts and lets the pipeline use different sandbox types and model providers.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/google/mantis/mantis-configure
Any agent
npx skills add google/mantis --skill mantis-configure
Clone the repo
git clone --depth 1 https://github.com/google/mantis

Made for: Claude Code, Codex.

Per session 77 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,706 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00077 $0.01706
Opus 5 $0.00039 $0.00853
Sonnet 5 $0.00015 $0.00341
Haiku 4.5 $0.00008 $0.00171

Measured 2d ago against content hash 16e191ace4eb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

mantis-configure 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 2d 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.

reference/skills/mantis-configure/SKILL.md · 175 lines

How it starts

The opening of the file, as written. The whole thing — 175 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Pipeline Configurator (/mantis-configure)

System Goal

Environment and Model Configurator. Configures workflow.json with appropriate sandbox execution mechanisms, AI model providers, API endpoints, and credential bindings. Provides instantaneous preflight verification to guarantee that LLM credentials and sandbox isolation requirements are fully operational before launching security review campaigns.

Command Definition

  • Command: /mantis-configure
  • Description: Configures Mantis pipeline settings (sandboxes, models, credentials, preflight validation) in workflow.json.
  • Execution Command:
    # From reference/ directory:
    python3 scripts/configure.py [flags...]
    
    # From repository root:
    python3 reference/scripts/configure.py [flags...]
    
  • CLI Options:
    • --sandbox / -s: Sandbox mechanism (static-only, gvisor, microsandbox, gce).
    • --model / -m: Default LLM model (e.g. gemini-3.7-flash, vertex_ai/claude-opus-5, vertex_ai/zai_org/glm-5.2-maas, openai/{MODEL_ID}).
    • --api-base: Custom endpoint URL for OpenAI-compatible LLM servers (e.g. http://localhost:8000/v1).
    • --reasoning-effort: Reasoning effort level (low, medium, high).
    • --timeout: LLM request timeout in seconds.
    • --project / -p: GCP Project ID (for GCE sandbox or Vertex AI routing).
    • --zone / -z: GCP Zone (e.g. us-central1-b).
    • --image / -i: Sandbox image name (e.g. mantis-sandbox-image or mantis-sandbox:latest).
    • --subnet: GCE Subnet name (e.g. mantis-isolated-subnet).
    • --workdir: Sandbox guest workdir (default: /workspace).
    • --workflow / -w: Path to workflow.json (defaults to auto-discovery).
    • --db / -d: Path to SQLite knowledge database (default: knowledge.db).
    • --auto: Auto-detects host capabilities and configures optimal settings automatically.
    • --save: Explicitly saves configuration changes to workflow.local.json.
    • --save-tracked / --global: Saves configuration changes directly to base workflow.json.
    • --interactive: Interactive step-by-step terminal wizard.
    • --test / --preflight: Executes fast (1-2s) validation tests verifying LLM reachability and sandbox readiness.
    • --show: Displays current configuration and diagnostic status.
    • --dry-run: Simulates configuration changes without modifying files.
    • --update-nodes: Updates all agent nodes in workflow.json to use the specified default model.
    • --json: Outputs configuration status and preflight diagnostics in JSON.

Read the full file on GitHub · 175 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. 2d ago First seen · 175 lines · 77 tokens per session scan A 16e191ace4eb

Subscribe to this mod's changes

mantis-configure is a skill published in the GitHub repository google/mantis (853 stars, last pushed 5d ago), licensed Apache-2.0. It adds 77 tokens to every session and 1,706 once invoked, about $0.0004 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 2026-08-30.