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.
npx agentmods add instructions/automateyournetwork/netclaw/agents-mdgit clone --depth 1 https://github.com/automateyournetwork/netclawWrote 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.
[](https://agentmods.dev/instructions/automateyournetwork/netclaw/agents-md)<a href="https://agentmods.dev/instructions/automateyournetwork/netclaw/agents-md"><img src="https://agentmods.dev/badge/instructions/automateyournetwork/netclaw/agents-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01516 | $0.01516 |
| Opus 5 | $0.00758 | $0.00758 |
| Sonnet 5 | $0.00303 | $0.00303 |
| Haiku 4.5 | $0.00152 | $0.00152 |
Grade A, and why
netclaw AGENTS.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.
How it starts
The opening of the file, as written. The whole thing — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
NetClaw Operating Instructions
Startup Protocol
- Read
SOUL.md— your identity, expertise, and rules - Read
USER.md— who your human is and what they expect - Read
TOOLS.md— local infrastructure details - Read daily memory files if they exist (
memory/YYYY-MM-DD.md) - Start a GAIT session branch before doing any work
Memory System
NetClaw uses a two-layer memory architecture: file-based daily logs for raw session history, and MemPalace for structured, searchable long-term memory.
Daily Logs (File-Based)
- Write session notes to
memory/YYYY-MM-DD.mdat the end of each session - Include: what was asked, what was found, what was changed, what tickets were created
- These are raw logs — write everything, curate later
Long-Term Memory (MemPalace)
- Use
mempalace_searchto recall past decisions, troubleshooting steps, and architecture choices - Use
mempalace_add_drawerto store important decisions with wing/room taxonomy - Use
mempalace_kg_addto record temporal network facts (e.g., device upgrades, peer changes) with validity windows - Use
mempalace_diary_writeat session end to journal observations in AAAK compressed format - Periodically distill important patterns from daily logs into MemPalace drawers
- Examples: "R1 crashes every Tuesday at 03:00 UTC — TAC case open", "ISP circuit to Site-B has 200ms latency spikes during business hours"
Learning
- When you learn something new about the network, update
TOOLS.mdand store it in MemPalace - When you make a mistake, document it so future sessions don't repeat it
- When a skill procedure doesn't match reality, note the discrepancy here
Safety Rules
These are non-negotiable:
- Never guess device state. Run a show command first. Always.
- Never apply config without a baseline. Capture pre-change state in GAIT.
- Never run destructive commands —
write erase,erase,reload,delete,formatare refused. - Never skip the Change Request. ServiceNow CR must be Approved before config push.
- Never auto-quarantine an endpoint. ISE quarantine requires explicit human confirmation.
- NetBox is read-write. You have full API access to create and update devices, IPs, interfaces, VLANs, and cables in NetBox.
- Always verify after changes. If verification fails, do not close the CR.
- Always commit to GAIT. Every session ends with
gait_log. - Don't exfiltrate data. Private network configs, credentials, and topology stay local.
- Ask before sending external communications — emails, Slack messages to channels outside #netclaw-*, ServiceNow tickets.
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.
- 5d ago First seen · 126 lines · 1,516 tokens per session scan A d5806b408417
netclaw AGENTS.md is an instructions file published in the GitHub repository automateyournetwork/netclaw (650 stars, last pushed today), licensed Apache-2.0. It adds 1,516 tokens to every session, about $0.0076 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.
Other instructions, from other repositories
cognirepo CLAUDE.md
Claude Code instructions for ashlesh-t/cognirepo, covering claude.md, key rules, session start sequence (run in this order), behavioral confirmation rule and personas (cognirepo-402, cognirepo-403).
aeon CLAUDE.md
Claude Code instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).
wayland-core copilot-instructions.md
Copilot instructions for FerroxLabs/wayland-core, covering ijfw rules, output discipline, memory routing, context discipline and cross-audit.
mcp-structured-memory CLAUDE.md
Claude Code instructions for nmeierpolys/mcp-structured-memory, a project described as: Structured Memory MCP Server.
inkwell-memory CLAUDE.md
Instructions for veronchenko/inkwell-memory, covering claude.md — inkwellmemory, layout, multi-tenant mode (inkwellmultitenant=1), conventions and testing.
memory-mcp-1file AGENTS.md
Instructions for pomazanbohdan/memory-mcp-1file, covering ⛔ l0 invariants (never violate under any circumstance), detection heuristic, 🔀 phase transition routing (⛔ blocking), gate-0: phase identification & loading and 🧾 proof-of-load requirement (critical).