skillsFirst-partyReview first Safety · Privacy ·

MCP Server Authoring Security Capability Pack Skill

Expert MCP capability skill for secure server authoring, tool schema discipline, auth boundaries, and adversarial prompt hardening.

by JSONbored·added 2026-04-10·

Claude CodeCodexWindsurfGeminiCursorCLI

HarnessClaude CodeCodexWindsurfGeminiCursorCLI

Level:expertType:capability-packVerified:validated

Install

Source

curl -L https://heyclau.de/downloads/skills/mcp-server-authoring-security-capability-pack.zip -o mcp-server-authoring-security-capability-pack.zip && unzip -o mcp-server-authoring-security-capability-pack.zip -d ./mcp-server-authoring-security-capability-pack

Readiness

TrustTrusted
Sourcefirst-party
Safety notesMissing
ReviewedYes

Documentation Source repository Registry JSON · LLM text

Review first — review before installing

Open the source and read safety notes before installing.

Prerequisites

MCP server implementation or design draft
Tool inventory and expected consumers
Logging/alerting sink for security events

Schema details

Install type: package
Reading time: 9 min
Difficulty score: 90
Troubleshooting: Yes
Breaking changes: No

Package metadata

Download URL: /downloads/skills/mcp-server-authoring-security-capability-pack.zip
Package verified: Yes
SHA-256: 88cfe7f7cbd688806926439fcef04929c691deaaa01909a87d5658bbeb8c28bb

Skill and platform metadata

Skill type: capability-pack
Skill level: expert
Verification: validated
Verified at: 2026-04-10

Retrieval sources

https://modelcontextprotocol.io/specification/2025-06-18/basic/security_best_practiceshttps://modelcontextprotocol.io/specification/2025-06-18https://modelcontextprotocol.io/docs/

Tested platforms

ClaudeCodexOpenClawCursorWindsurfGemini

Platform	Support	Install path
claude-code	Native	.claude/skills/<skill-name>/SKILL.md
codex	Native	.agents/skills/<skill-name>/SKILL.md
windsurf	Native	.windsurf/skills/<skill-name>/SKILL.md
gemini	Native	.gemini/skills/<skill-name>/SKILL.md or .agents/skills/<skill-name>/SKILL.md
cursor	Adapter	.cursor/rules/<skill-name>.mdc
cli	Manual	AGENTS.md or tool-specific context file

Full copyable content

# Trigger
"Apply the MCP server authoring security capability pack to this server."

# Required output
1) Threat model and trust boundaries
2) Tool schema and validation contracts
3) Auth and permission architecture
4) Abuse and incident response controls

About this resource

Knowledge Freshness

This capability pack is pinned to documentation verified on 2026-04-10. When upstream docs change, refresh endpoint contracts, examples, and constraints before using this skill for production changes.

Retrieval Sources

Always prefer direct retrieval from official docs/API references over model memory for limits, endpoint signatures, and behavior guarantees.

Core Workflow

Confirm target version/runtime and pull latest official docs for the task scope.
Build an execution plan with explicit read-only discovery before any mutation.
Validate contracts, permissions, and safety constraints before applying changes.
Execute with deterministic checkpoints and rollback criteria.
Produce a verification report with evidence, caveats, and next actions.

Overview

This capability pack teaches agents how to design MCP servers that remain safe under real adversarial conditions. It emphasizes least privilege, explicit contracts, and secure default behavior.

Capability Scope

Threat modeling for tool invocation flows
Input validation and schema hardening
AuthN/AuthZ boundaries for tools and resources
Prompt-injection and data exfiltration defenses
Audit logging and incident response readiness

Compatibility

Native

Claude Code / Claude: native skill usage via SKILL.md.
Codex/OpenAI workflows: compatible with Agent Skills-style SKILL.md content as reusable workflow instructions.

Manual Adaptation

Gemini CLI: native skill usage via .gemini/skills/<skill-name>/SKILL.md or .agents/skills/<skill-name>/SKILL.md where supported.
Cursor: use the generated .cursor/rules/*.mdc adapter for project rules.
OpenClaw and similar agents: use the same skill content as a reusable prompt/workflow file when native skill import is unavailable.

Production Rules

Validate all tool input before execution.
Scope each tool to minimal permissions.
Deny by default for unsafe/unknown operations.
Preserve immutable audit trails for sensitive invocations.

Troubleshooting

Issue: Tool abuse through crafted prompts
Fix: Add policy layer between prompt interpretation and tool invocation.

Issue: Schema drift breaks clients
Fix: Enforce contract tests and backward compatibility checks.

Issue: Difficult to investigate incidents
Fix: Add structured event logs with actor/tool/input/result correlation.

Output Contract

Provide an implementation plan ordered by risk and dependency.
Provide exact production-ready config/commands with no placeholders.
Call out secrets, permissions, and least-privilege requirements.
Include rollback and recovery guidance for each risky step.

Validation Checklist

Verify all referenced docs/versions before applying changes.
Run regression checks for core user flow and error paths.
Confirm observability/logging is enabled for changed components.
Confirm security controls (auth, rate limits, input validation) still pass.
Record final known limitations and follow-up actions.

Content outline

Knowledge Freshness
Retrieval Sources
Core Workflow
Overview
Capability Scope
Compatibility
Native
Manual Adaptation
Production Rules
Troubleshooting
Output Contract
Validation Checklist

#mcp#security#server-authoring#tools#capability-pack

Source citations

Signals

Loading live community signals…