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 skills add cosmicstack-labs/mercury-agent-skills --skill ai-agent-designgit clone --depth 1 https://github.com/cosmicstack-labs/mercury-agent-skillsWrote 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/skills/cosmicstack-labs/mercury-agent-skills/ai-agent-design)<a href="https://agentmods.dev/skills/cosmicstack-labs/mercury-agent-skills/ai-agent-design"><img src="https://agentmods.dev/badge/skills/cosmicstack-labs/mercury-agent-skills/ai-agent-design/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/cosmicstack-labs/mercury-agent-skills/ai-agent-design"><img src="https://agentmods.dev/badge/skills/cosmicstack-labs/mercury-agent-skills/ai-agent-design.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 5 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high YARA Match · line 3 YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).Fix: Remove offensive tool references and exploit code. Legitimate agent skills should not contain penetration testing tools, exploit frameworks, or reconnaissance utilities.
- high Prompt Injection · line 524 This pattern attempts to override system instructions or ignore safety constraints. Without LLM analysis, manual review is recommended.Fix: Remove or rewrite any text that instructs the agent to ignore prompts, override safety rules, or trust unverified content. Ensure skill content cannot be injected to alter agent behavior.
- high System Prompt Leakage · line 544 Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.Fix: Remove any instructions that reveal, print, or output system prompts or internal rules. System instructions should never be exposed to end users.
- high Prompt Injection · line 643 This pattern attempts to override system instructions or ignore safety constraints. Without LLM analysis, manual review is recommended.Fix: Remove or rewrite any text that instructs the agent to ignore prompts, override safety rules, or trust unverified content. Ensure skill content cannot be injected to alter agent behavior.
- medium Excessive Agency · line 568 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 5.1 | $0.00048 | $0.05231 |
| Opus 5 | $0.00024 | $0.02616 |
| Sonnet 5 | $0.00010 | $0.01046 |
| Haiku 4.5 | $0.00005 | $0.00523 |
Grade B, and why
ai-agent-design scanned grade B with 1 finding 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 11d 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.
Instruction-override phrasingmediumPrompt injection
Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.
r"ignore all previous instructions", Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 670 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI Agent Design
Core Principles
1. Agents Are Tools, Not Teammates
An AI agent is a system that uses an LLM to reason and take actions. It is not a person — it has no goals, desires, or understanding. Design agents as tools with clear boundaries, not as autonomous collaborators.
2. Autonomy is a Spectrum
Full autonomy is rarely the goal. The best agents operate on a spectrum: more human oversight for critical actions, more autonomy for routine tasks. Design for the level of autonomy that matches the risk.
3. Cache Everything, Guess Nothing
Agents have no memory between calls unless you design it. Every interaction, tool result, and decision must be explicitly stored and retrieved. Assume the agent remembers nothing unless you program it to.
4. Fail Predictably
Every agent will fail. The question is how it fails. Design for graceful degradation: when uncertain, ask for help. When stuck, escalate. When broken, stop safely.
5. Safety First, Speed Second
A fast agent that takes unauthorized actions is worse than a slow agent that double-checks. Build guardrails before building features.
Agent Maturity Model
| Level | Name | Characteristics | Tool Use | Memory | Autonomy |
|---|---|---|---|---|---|
| L1 | Reactive | Single-turn, no context retention, deterministic responses | None or hardcoded | None | None |
| L2 | Scripted | Pre-defined workflows, conditional branching, template-based | Basic function calls with fixed signatures | Session-only (ephemeral) | Low — requires human confirmation |
| L3 | Tool-Using | Dynamic tool selection, structured function calling, error handling | Multiple tools, runtime discovery | Short-term (conversation history) | Medium — executes routine tasks autonomously |
| L4 | Memory-Augmented | Long-term memory, learns from past interactions, personalization | Complex tools with parameter binding | Long-term + episodic (vector stores, databases) | High — manages complex workflows |
| L5 | Autonomous Orchestrator | Multi-agent coordination, dynamic planning, self-correction, meta-cognition | Tool composition, tool creation, delegation | Semantic + episodic (knowledge graphs, RAG) | Full — handles novel situations independently |
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.
- 11d ago First seen · 670 lines · 48 tokens per session scan B 37276a25514e
ai-agent-design is a skill published in the GitHub repository cosmicstack-labs/mercury-agent-skills (470 stars, last pushed 16d ago), licensed MIT. It adds 48 tokens to every session and 5,231 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (instruction-override phrasing). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
context-engineering
Optimizes agent context setup. Use when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and context for a project.
comet-memory
A review step for deciding whether information should become durable personal memory. It can keep, update, forget, or skip memory candidates based on bounded evidence.
recall-memory
Recall relevant long-term memories on demand. Given a topic or question, judges relevance from pre-loaded metadata, loads only relevant files, and returns a concise summary to the main agent.
agent-expert-creation
Create specialized agent experts with pre-loaded domain knowledge using the Act-Learn-Reuse pattern. Use when building domain-specific agents that maintain mental models via expertise files and self-improve prompts.
relevance-coarse-filter
Cheap, high-recall first-pass filter that removes obvious junk from a detector candidate pool before expensive story-origin research and PR judgment. Decides keep, monitoronly, or reject — never ranks, writes angles, verifies dates, or decides whether to pitch.
self-improve
Extract lessons from the current session, or sweep the project's past sessions when asked, and route them to the appropriate knowledge layer (project AGENTS.md, auto memory, existing skills, or new skills). Use when the user asks to "self-improve", "distill this session", "distill past sessions", "sweep past…