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/syntax753/synkb/claude-mdgit clone --depth 1 https://github.com/Syntax753/synkbWhat 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 | $0.00983 | $0.00983 |
| Opus 5 | $0.00491 | $0.00491 |
| Sonnet 5 | $0.00197 | $0.00197 |
| Haiku 4.5 | $0.00098 | $0.00098 |
Grade A, and why
synkb 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 yesterday.
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 — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Guidance for Claude Code working in this repository.
This file is deliberately short. It carries only what an agent needs that is not already written for humans, and points at the human docs for everything else. When guidance would apply equally to a human contributor, it belongs in CONTRIBUTING.md — not here. Two copies of one convention drift, and the drift is invisible until someone acts on the stale one.
Read these first
| Question | File |
|---|---|
| What is this project, how do I run it? | README.md |
| Service conventions, shared-py, dev workflow, commit style | CONTRIBUTING.md |
| What to build next, and what "done" means | design/implementation-plan.md |
| Why is it like that? | design/decision-log.md |
| Architecture and the demo→production mapping | design/hld.md |
| UI labels and flows (source of truth) | docs/user-guide.md |
| What is a deliberate demo weakness vs a bug | SECURITY.md |
Rules that are easy to get wrong
- Work is scoped by work package. Locate the WP in
design/implementation-plan.mdand treat its acceptance criteria as the contract. Don't widen scope; if something belongs to a later WP, leave it. - Record non-obvious decisions in
design/decision-log.md, newest at the top. The log is append-only: correct the record by adding an entry, never by editing history. - Documentation ships in the same commit as the change. A change without its doc
update is incomplete — including
docs/user-guide.mdwhen a UI label or flow moves. - Never hand-edit
apps/web/lib/api/schema.ts. It is generated from core-api's OpenAPI. Runmake typesand commit the result. - Unit tests are not sufficient for a mutation. vitest stubs
fetchand skips core-api's proxy entirely, so a broken gateway path passes every unit test. Exercise new mutations against a livemake demostack before calling the work done. DEMO_MODEbranches are deliberate, not bugs to clean up. Production flips the flag and real auth takes over.STUB:tools and "good enough for now" comments are intentional. They mark honest boundaries. Don't tidy them away — if a stub becomes real, delete the marker as part of that work.- Never commit machine-local paths, personal contact details, or secrets. The repo
is public. The maintainer's only published contact route is LinkedIn
peterturnerlondon; GitHub isSyntax753.
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.
- yesterday First seen · 75 lines · 983 tokens per session scan A b6ae0557c040
synkb CLAUDE.md is an instructions file published in the GitHub repository Syntax753/synkb (0 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 983 tokens to every session, about $0.0049 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 2026-08-31.
Other instructions, from other repositories
ai-bvf CLAUDE.md
Claude Code instructions for Craig-Horton/ai-bvf, covering working with craig, writing, design and repo facts.
martin-loop AGENTS.md
Instructions for Keesan12/martin-loop, covering martinloop agent operating rules, repository authority, mandatory internal-first sequence, prohibited actions and required internal health proof.
CORE CLAUDE.md
Instructions for DariuszNewecki/CORE, covering claude.md — core, what core is, and what you are, source layout, how to work in this repo and governed and prohibited surfaces.
open-ace CLAUDE.md
Instructions for open-ace/open-ace, covering claude.md, where process documents go, pushing, test placement and ci semantics and schema snapshots (schema-sync ci).
gate CLAUDE.md
Instructions for GaaraZhu/gate, covering gate, notes, repository structure, build and test commands and before every commit.
orloj AGENTS.md
Instructions for OrlojHQ/orloj, covering orloj agent instructions, must-follow sync rules and working style.