# The Hounds — navi-agents

> The executable harness for The Hounds — a local console that runs the exposure-management agent pack over Tenable navi.

- Type: Playbook
- Author: packetchaos
- License: MIT
- Trust tier: contributed (passed automated validation and a baseline maintainer review)
- Last reviewed: 2026-07-17
- Source repository: https://github.com/packetchaos/the-hounds-repo
- Repo state: not archived
- Last push: 2026-09-17T00:15:07Z
- Stars: 1 (a popularity signal, not a safety signal)
- Forks: 1 (a popularity signal, not a safety signal)
- Open issues: 0
- Language: HTML
- Health checked at: 2026-11-05T20:35:15.821Z
- Integrations: Tenable, Anthropic
- Domains: vulnerability-management, platform-operations
- Tags: tenable, exposure-management, navi, playbook, python, console, vulnerability-management
- Date added: 2026-07-15
- Playbook type: standard
- Agent chain:
  1. Laelaps (info), Finds and tags CISA KEV (known exploited) exposure.
  2. Certania (info), Tracks certificate expiry and weak crypto.
  3. Heimdall (info), Assesses post-quantum readiness.
  4. Fenrir (info), Chains signals into ranked attack paths (foothold to crown jewel).
  5. Cerberus (info), Confidence-scored IoT / OT / embedded device discovery.
  6. Pythia (info), Discovers and governs AI/ML inventory across five sources.
  7. Atlas (info), Establishes asset ownership.
  8. Mimir (info), Software inventory.
  9. Charon (info), Flags end-of-life / unsupported software.
  10. Anubis (info), Calibrates Asset Criticality Rating (ACR).
  11. Chronos (info), Scan health monitoring.
  12. Sirius (info), Agent group analysis.
  13. Garmr (info), Tag removal and cleanup.
  14. Orthrus (info), Maps findings to MITRE ATT&CK.
  15. Argus (info), Custom application discovery.
  16. Argos (info), Single-asset deep-dive.
  17. Sphinx (info), 'On the Scent' environment overview.
  18. Covenant (info), Enforces the AI Contract governance policy.
- Listing page: https://exchange.tenable.com/playbooks/the-hounds-navi-agents
- Source file: https://github.com/tenable/cyberagents-exchange/blob/main/playbooks/the-hounds-navi-agents.md

The Hounds — navi-agents is the executable harness for the pack: a local console
(`navi-agents/`) that runs the exposure-management hounds over
[Tenable **navi**](https://github.com/packetchaos/navi) (`navi.db` + the Tenable API) and
surfaces the results in the browser. It runs on a stock Python install — the default server
needs no third-party packages.

## What it does

It executes the same pack of specialists as a program rather than a skill: CISA KEV
(Laelaps), attack paths (Fenrir), IoT/OT (Cerberus), AI inventory (Pythia), certificates
(Certania), EOL software (Charon), ACR calibration (Anubis), and the rest. Each hound
hunts one kind of exposure and proposes navi tag/ACR writes for human approval; writes are
disabled by default (`NAVI_ALLOW_WRITES=0`).

## How it works

Launch with `python3 run.py` (zero-dependency stdlib server, or optional FastAPI/Flask).
On first run it builds a bundled sample database so you can evaluate it with no setup; for
production, point `NAVI_DB_PATH` at navi's real `navi.db`. The `core/` engine provides the
shared detection, discovery, health, MITRE, and EOL logic, and `core/agents/` holds the
individual hound implementations. Every write stays grounded in real navi data and gated
behind explicit confirmation.

This is the Markdown version of one of 140 listings on CyberAgents Exchange. The same listing as JSON: https://exchange.tenable.com/api/playbooks/the-hounds-navi-agents.json. The full index and how to read it: https://exchange.tenable.com/llms.txt