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/jmanhype/speckit/agents-mdgit clone --depth 1 https://github.com/jmanhype/speckitWhat 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 | $0.04083 | $0.04083 |
| Opus 5 | $0.02041 | $0.02041 |
| Sonnet 5 | $0.00817 | $0.00817 |
| Haiku 4.5 | $0.00408 | $0.00408 |
Grade A, and why
speckit 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 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.
How it starts
The opening of the file, as written. The whole thing — 531 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md - Instructions for AI Coding Agents
This file contains workflow rules and guidelines for AI agents (Claude Code, GitHub Copilot, Cursor, Amp, etc.) working in repositories that use Spec Kit + Beads.
Core Philosophy
Spec Kit provides structure (WHAT/WHY/HOW). Beads provides memory (persistent task graph that survives context limits).
Together they solve:
- ✅ Structured specification-driven development
- ✅ Long-term memory across sessions
- ✅ Dependency tracking that doesn't disappear
- ✅ Work discovery and prioritization
Pivotal Methodology Alignment
Spec Kit + Beads implements practices from Pivotal Labs (now VMware Tanzu Labs):
| Pivotal Practice | Spec Kit + Beads Implementation |
|---|---|
| TDD | test-gate.sh enforces 100% test pass after every edit |
| User Stories | spec.md with P1/P2/P3 priorities, [US#] references |
| Story Types | Beads: --type epic/task/bug, labels for chores |
| Story States | Beads: --status todo/in-progress/done/blocked |
| Acceptance Criteria | spec.md ## Acceptance Criteria → Beads epic description |
| IPM (Planning) | /speckit.specify → /speckit.plan → /speckit.tasks |
| Velocity | Beads tracks task completion; bd ready shows unblocked work |
| Dependencies | bd dep add creates blocking relationships (P0 → P1 → P2) |
Pivotal-Style Epic Format
Beads epics created by Spec Kit include:
PROBLEM STATEMENT:
[From spec.md ## Problem Statement]
BUSINESS VALUE:
[From spec.md ## Business Value]
ARCHITECTURAL VISION:
[From plan.md ## Architectural Vision]
INTEGRATION TESTS:
[From spec.md ## Integration Tests]
Acceptance Criteria:
[From spec.md ## Acceptance Criteria]
Blocks (N):
← TASK-1: P0 task description
← TASK-2: P1 task description
...
Priority Levels (Pivotal-Aligned)
| Priority | Pivotal Equivalent | Description |
|---|---|---|
| P0 | Urgent/Blocker | Critical path, blocks everything else |
| P1 | High | MVP features, must-have |
| P2 | Medium | Should-have, important but not blocking |
| P3 | Low | Nice-to-have, future enhancements |
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.
- 2d ago First seen · 531 lines · 4,083 tokens per session scan A 523d803eb270
speckit AGENTS.md is an instructions file published in the GitHub repository jmanhype/speckit (26 stars, last pushed 5mo ago), licensed MIT. It adds 4,083 tokens to every session, about $0.0204 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 2026-09-01.
Other instructions, from other repositories
scaffold AGENTS.md
Instructions for zigrivers/scaffold, covering agent instructions, quick reference, scaffold releases and landing the plane (session completion).
scaffold CLAUDE.md
Instructions for zigrivers/scaffold, covering claude.md, core principles, project overview, structure and source of truth.
claude-code-templates CLAUDE.md
Claude Code instructions for Justdvp/claude-code-templates, covering claude.md, project overview, development commands, package management and application commands.
codexspec CLAUDE.md
Claude Code instructions for Zts0hg/codexspec, covering claude.md - codexspec development guide, project overview, purpose, technology stack and architecture.
flow AGENTS.md
Instructions for cofin/flow: Flow is a toolkit for context-driven development. It has no flow executable; invoke the Flow skill or a /flow: command.
trellis AGENTS.md
Instructions for craigcossairt/trellis, covering agents.md , project, what i need from agents, current state - source of truth pointers and tech stack.