What is agent-knowledge-cards AGENTS.md?

Repository instructions for agent-knowledge-cards, including its memory system, host adapters, command-line tools, MCP server, and evaluation process. An AGENTS.md file gives coding agents rules for working in a project.

Guide agents working on the repository, including where to edit code, how memory and adapters are separated, and how to run evaluation checks.

It gives agents clear project-specific rules about structure, writing style, commands, and validation, reducing inconsistent changes.

agent-knowledge-cards AGENTS.md — Instructions file for Codex and OpenCode by manojbajaj95: Repository instructions for agent-knowledge-card

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 instructions/manojbajaj95/agent-knowledge-cards/agents-md

Clone the repo

git clone --depth 1 https://github.com/manojbajaj95/agent-knowledge-cards

Made for: Codex, OpenCode.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/manojbajaj95/agent-knowledge-cards/agents-md.svg)](https://agentmods.dev/instructions/manojbajaj95/agent-knowledge-cards/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/manojbajaj95/agent-knowledge-cards/agents-md"><img src="https://agentmods.dev/badge/instructions/manojbajaj95/agent-knowledge-cards/agents-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 879 This file is loaded in full into every session.

When invoked 879 The same file — it is already loaded in full.

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.

Model	Per session	Once invoked
Fable 5	$0.00879	$0.00879
Opus 5	$0.00439	$0.00439
Sonnet 5	$0.00176	$0.00176
Haiku 4.5	$0.00088	$0.00088

Measured today against content hash 7f5f592ada20, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agent-knowledge-cards 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 today.

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.

AGENTS.md · 59 lines

How it starts

The opening of the file, as written. The whole thing — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Always talk in ASD-STE100 Simplified Technical English. Also follow Zinsser's four principles of quality writing:

Simplicity
Brevity
Clarity
Humanity

Conventions for agents working in this repo.

Layout

Two boxes: memory (save, store, retrieve) and harness (fetch + reflect). Adapters wrap host envelopes only.

src/memory/ — L1 card internals (ingest, retrieve, storage, markdown) + persist ops (ops.ts). Host-agnostic.
src/harness/ — session fetch (fetchCards) + reflect follow-up (reflectFollowup, REFLECT.md). Not a host SDK.
src/cli/ — thin CLI over memory (status, query, propose, update, delete, mcp) plus install (adapter wiring). init is hidden; propose creates dirs.
src/adapters/ — host hook envelopes (Claude Code / Cursor / Codex) and the Pi extension. Do not import src/memory.
src/mcp/ — MCP stdio server (@modelcontextprotocol/sdk); tool logic in tools.ts.
eval/ — Harbor A/B with/without knowcards (eval:prepare / eval:run / eval:compare). Primary validation for this slice (manual; not CI). Sequential same-repo runs are named only.
tests/ — offline checks for retrieve/fetch and the eval pipeline. No separate unit-test suite beyond that.
dist/ — build output for npm / npx (do not edit; emit with bun run build).

On-disk cards

Filesystem-first: cards are local markdown only (.agents/knowledge_cards/<notebook-id>/*.md). No DB backend yet. propose creates notebook dirs as needed (init is hidden; not a required step). Process start loads the full library into memory via openLibrary.

Commands

bun install
pre-commit install       # once per clone
bun test                 # eval pipeline offline checks only
bun run typecheck
bun run lint
bun run build
bun run knowcards install cursor|claude-code|codex|pi
bun run knowcards status|query|propose|update|delete|mcp
bun run eval:prepare     # Harbor — manual
bun run eval:run -- --task pytest-dev__pytest-10051 --agent oracle
bun run eval:run -- --task pytest-dev__pytest-10051  # [email protected] + openai/gpt-5.6-luna
bun run eval:run  # four SWE-bench Verified tasks

Read the full file on GitHub · 59 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.

today Changed · +18 tokens per session 7f5f592ada20
5d ago First seen · 59 lines · 861 tokens per session scan A 05239ef128e7

Subscribe to this mod's changes

agent-knowledge-cards AGENTS.md is an instructions file published in the GitHub repository manojbajaj95/agent-knowledge-cards (4 stars, last pushed 5d ago), licensed MIT. It adds 879 tokens to every session, about $0.0044 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.

Building agents Memory and context TypeScript agents ai-agents ai-memory long-term-memory

hindsight CLAUDE.md

Claude Code instructions for vectorize-io/hindsight, covering claude.md, project overview, development commands, local development (api + ui) and start both api server and control plane ui.

vectorize-io/hindsight · 4,858 tokens

hindsight AGENTS.md

AGENTS.md instructions for vectorize-io/hindsight: See CLAUDE.md for project documentation and coding conventions.

vectorize-io/hindsight · 25 tokens

InvOS AGENTS.md

AGENTS.md instructions for felipenalves/InvOS, covering agents.md — invos (sistema operacional do negócio), contexto do negócio, primeira vez, fluxo de trabalho and aprender com correções.

felipenalves/InvOS · 1,340 tokens

memory-lancedb-mcp CLAUDE.md

Claude Code instructions for cablate/memory-lancedb-mcp, covering claude.md, project principles, git workflow, development commands and architecture.

cablate/memory-lancedb-mcp · 1,019 tokens

opengeni CLAUDE.md

Claude Code instructions for Cloudgeni-ai/opengeni, a project described as: Open, self-hostable agentic runtime for organizations — durable, replayable agent sessions, human approvals, governed credentials and memory, running in managed sandboxes or on your own hardware. Apache-2.0.

Cloudgeni-ai/opengeni · 3 tokens

InvOS CLAUDE.md

Claude Code instructions for felipenalves/InvOS, a project described as: INVOS — memória pro seu agente de IA. Para de reexplicar quem você é em toda sessão. Multi-harness (Cursor, Claude, Codex, Grok…).

felipenalves/InvOS · 5 tokens