MCP App StoreMCP App Store
from cskill-agents36
Optimizes error handling for token limit failures by delaying reporting until recovery loops complete, providing cleaner tool diagnostics.
This skill improves the user experience when an AI agent hits the max_output_tokens limit during a streaming multi-tool response. Instead of immediately crashing or leaking a raw API error, it suppresses the error while the system's internal recovery mechanisms attempt to resolve the issue.
Use this skill when implementing tool-orchestration loops where high-token output is common. It is specifically designed for scenarios where a recovery loop should finish its attempt before the user is notified of a failure.
tool_result placeholders to keep session context intact.Primarily designed for Claude Code and other agents utilizing similar streaming recovery loops and tool-calling architectures.
This skill has not been reviewed by our automated audit pipeline yet.
Capability Based Spinner Glyph Selection
Dynamically select spinner glyphs and colors based on terminal capabilities, motion preferences, and stall states.
No-Op Empty Skill Delta Return
Prevents the emission of empty update payloads when incremental skill deltas contain no new items, reducing stream noise.
Agent-Specific MCP Server Initialization
Safely initializes and tears down MCP servers defined in an agent's frontmatter without leaking clients into the parent session.
Websocket Client Creator
Generates and guides websocket client integrations: configuration, code samples, best practices, and validation for production-ready websocket clients.
Canary — Post-Deploy Visual Monitor
Run a short post-deploy monitor that captures screenshots, checks console errors, and compares performance against baselines to detect regressions and page fail
API and Interface Design
Guides design of stable, predictable APIs and interfaces — contract-first patterns, error semantics, pagination, and naming conventions for robust integrations.
API Development
Guidelines and patterns for building robust REST APIs: correct HTTP status codes, RFC 7807 error responses, validation, sanitization, rate limiting, and central
MCAF Non-Functional Requirements (NFR)
Turn vague quality goals into explicit, testable non-functional requirements (accessibility, reliability, performance, scalability, maintainability, compliance)
TaskEither Quick Reference (fp-ts)
Concise cheat-sheet for fp-ts TaskEither: patterns for typed async error handling, composing Promises, and recoverable pipelines.
Decomposition & Reconstruction
Break complex systems into components, map relationships, and reconstruct optimized configurations to identify bottlenecks, failure points, and redesign opportu
System Design
Structured, end-to-end guidance for designing scalable, reliable systems: requirements, capacity estimation, architecture, and failure-mode planning.