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/ssdeanx/ssd-ai/agents-mdgit clone --depth 1 https://github.com/ssdeanx/ssd-aiWrote 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/ssdeanx/ssd-ai/agents-md)<a href="https://agentmods.dev/instructions/ssdeanx/ssd-ai/agents-md"><img src="https://agentmods.dev/badge/instructions/ssdeanx/ssd-ai/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 | $0.03300 | $0.03300 |
| Opus 5 | $0.01650 | $0.01650 |
| Sonnet 5 | $0.00660 | $0.00660 |
| Haiku 4.5 | $0.00330 | $0.00330 |
Grade A, and why
ssd-ai 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 4d 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 — 266 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Guidance for automated coding agents
Build / Run / Test commands
- Build:
npm run build(runstsc) - Dev (stdio):
npm run dev:stdioornpm run dev(use CLI flags:--transport=stdio) - Start HTTP server:
npm run start:http(script passes CLI flags--transport=http --port=3000 --hostname=localhost) - Run all tests:
npm run test(uses Vitest) - Run single test file:
npx vitest run <path/to/testfile>ornpm run test -- <path/to/testfile>(example:npx vitest run src/__tests__/taskManager.test.ts) - Watch tests:
npm run test:watch; UI:npm run test:ui; Coverage:npm run test:coverage - Type-check only:
tsc --noEmit - Lint/format:
npx biome checkandnpx biome format(usenpx biome format --applyto rewrite files)
Code style & conventions (follow existing code style)
- Formatting: use Biome as source of truth (single quotes, semicolons, consistent spacing).
- Imports: external packages first, blank line, internal modules, then relative imports; keep grouped and minimal deep relative paths.
- Filenames: use camelCase for
.tsfiles (match existing code likeMemoryManager.ts). - Naming:
camelCasefor variables/functions,PascalCasefor types/classes/interfaces (noIprefix). - Types: prefer explicit return types on exported functions, avoid
any— useunknownfor untrusted input and narrow/cast. - Error handling: catch errors, prefer domain errors (e.g.
McpError) and rethrow typed errors; useinstanceof Errorwhen reading message. - Logging: avoid
console.login library code; useconsole.erroronly for entrypoints and failures.
Repo rules
- Copilot: see
.github/copilot-instructions.md(present, currently empty) — respect project privacy and review suggestions. - Cursor rules: no
.cursoror.cursorrulesfound in repository.
If you need me to run tests, lint, or adjust these rules, tell me which command to run.
src/ — Source layout & conventions
Overview
- src/ is the main TypeScript source tree; entry point is src/index.ts. It contains core libraries, helpers, transports, shared types, and unit tests.
Directory breakdown
- src/index.ts — public exports/entrypoint
- src/lib/ — core domain libraries & services (ContextCompressor.ts, MemoryManager.ts, ProjectCache.ts, PythonParser.ts, taskManager.ts)
- src/tools/browser/ — Browser automation and inspection tools (browserUtils.ts: general browser utilities, inspectNetworkRequests.ts: monitors HTTP requests, monitorConsoleLogs.ts: captures console output for debugging)
- src/tools/convention/ — Code quality and convention enforcement (analyzeComplexity.ts: measures code complexity, applyQualityRules.ts: applies coding standards, checkCouplingCohesion.ts: analyzes module dependencies, getCodingGuide.ts: retrieves coding guidelines, suggestImprovements.ts: proposes code enhancements, validateCodeQuality.ts: checks code against quality metrics)
- src/tools/memory/ — Session and context memory management (autoSaveContext.ts: automatically saves session state, deleteMemory.ts: removes stored memories, listMemories.ts: displays saved memories, prioritizeMemory.ts: ranks memories by importance, recallMemory.ts: retrieves specific memories, restoreSessionContext.ts: restores previous sessions, saveMemory.ts: stores new memories, searchMemories.ts: queries stored memories, startSession.ts: initializes memory sessions, updateMemory.ts: modifies existing memories)
- src/tools/planning/ — Project planning and requirements analysis (analyzeRequirements.ts: breaks down project needs, createUserStories.ts: generates user stories, featureRoadmap.ts: creates development roadmaps, generatePrd.ts: produces product requirement documents)
- src/tools/prompt/ — AI prompt analysis and enhancement tools (analyzePrompt.ts: evaluates prompt effectiveness, browserDevelopment.ts: aids in browser-related prompts, codeConventions.ts: enforces prompt coding standards, codeReview.ts: assists with code review prompts, debugAssistant.ts: helps with debugging prompts, documentation.ts: improves documentation prompts, enhancePrompt.ts: refines prompts for better AI responses, enhancePromptGemini.ts: Gemini-specific prompt enhancements, memoryManagement.ts: manages prompt memory, projectPlanning.ts: supports planning prompts, promptEnhancement.ts: general prompt improvement, reasoningFramework.ts: applies reasoning to prompts, semanticAnalysis.ts: analyzes prompt semantics, sequentialThinking.ts: enables step-by-step prompt processing, specGeneration.ts: generates specifications from prompts, testing.ts: aids in testing prompts, timeUtilities.ts: handles time-related prompts, uiPreview.ts: previews UI from prompts)
- src/tools/reasoning/ — Reasoning framework application (applyReasoningFramework.ts: implements structured reasoning processes)
- src/tools/resources/ — Resource and capability management (capabilities.ts: defines system capabilities, readme.ts: handles README generation)
- src/tools/semantic/ — Code semantic analysis (findReferences.ts: locates code references, findSymbol.ts: searches for symbols, README.md: documentation for semantic tools)
- src/tools/thinking/ — Problem-solving and thinking aids (analyzeProblem.ts: breaks down problems, breakDownProblem.ts: decomposes complex issues, createThinkingChain.ts: builds reasoning chains, formatAsPlan.ts: structures thoughts as plans, stepByStepAnalysis.ts: provides incremental analysis, thinkAloudProcess.ts: simulates thinking aloud)
- src/tools/time/ — Time-related utilities (getCurrentTime.ts: retrieves current time)
- src/tools/ui/ — User interface preview tools (previewUiAscii.ts: renders UI in ASCII format)
- src/transports/ — transport implementations (http.ts)
- src/types/ — shared DTOs and type definitions (pagination.ts, prompt.ts, resource.ts, task.ts, tool.ts)
- src/tests/ — unit tests (taskManager.test.ts, pagination.test.ts)
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.
- 4d ago First seen · 266 lines · 3,300 tokens per session scan A 1dcea101d3ff
ssd-ai AGENTS.md is an instructions file published in the GitHub repository ssdeanx/ssd-ai (3 stars, last pushed 8mo ago), licensed MIT. It adds 3,300 tokens to every session, about $0.0165 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
Genesis-World-MCP AGENTS.md
AGENTS.md instructions for Gnidreve/Genesis-World-MCP, covering agents.md — cas genesisworld mcp, what this is, standing orders — read first, design pillars and source layout & conventions.
atelier-mcp copilot-instructions.md
Copilot instructions for anshrajore/atelier-mcp, covering run npm run sync-adapters to update, universal ui/ux rules, universal backend architecture rules and framework presets (available in skills/atelier/presets/).
atelier-mcp CLAUDE.md
Claude Code instructions for anshrajore/atelier-mcp, covering atelier quality gate adapter for claude code (claude.md), run npm run sync-adapters to update, universal ui/ux rules, universal backend architecture rules and framework presets (available in skills/atelier/presets/).
atelier-mcp AGENTS.md
AGENTS.md instructions for anshrajore/atelier-mcp, covering atelier quality gate adapter for agents.md, run npm run sync-adapters to update, universal ui/ux rules, universal backend architecture rules and framework presets (available in skills/atelier/presets/).
mcpscore AGENTS.md
Instructions for mcp-box/mcpscore, covering agent instructions for mcpscore and gotchas (learned the hard way).
rollbar-mcp-server CLAUDE.md
Claude Code instructions for rollbar/rollbar-mcp-server, covering claude.md, build and development commands and code style guidelines.