Yusuf Akçakaya
Community@fusuyfusuy · Istanbul, TR
Yusuf Akçakaya publishes agent-governance skills for multi-agent coordination, AST code intelligence, architectural review, and technical-debt tracking.
Agent Skills by Yusuf Akçakaya
Showing 11 vetted skills indexed across 1 GitHub repositories.
fusuyfusuyadd-skills
Enables shared skill suites for a single project via symlinks.
fusuyfusuyswarm-governance
Coordinates multi-agent subagent teams with role tiers, isolation rules, and structured handoff contracts.
fusuyfusuymimori
Provides AST-based symbol search, context slicing, and blast-radius analysis for codebases.
fusuyfusuyboundary-review
Audits codebases along subsystem boundaries and cross-boundary seams with scored review reports.
fusuyfusuypantheon
Selects an engineering ancestor persona and emits a one-line doctrine verdict for design decisions.
fusuyfusuyfast-iterate
Implements rapid vertical-slice coding with smoke-first TDD and minimal-diff verification loops.
fusuyfusuyakatsuki
Query and mutate a Markdown knowledge vault via CLI and MCP stdio tools.
fusuyfusuyadd-mcp-server
Enables optional MCP servers from a shared library for a single project.
fusuyfusuyarch-inquisitor
Interrogates software architecture designs to produce implementation-grade specifications with verified seams.
fusuyfusuysummarizer
Generates recency-weighted session summaries from agent conversation transcripts via a subagent.
fusuyfusuyponytail-debt
Harvest ponytail comment markers into a tracked technical debt ledger.
Frequently Asked Questions About Yusuf Akçakaya
FAQPage SchemaWhat tasks can I accomplish with Yusuf Akçakaya's skills?▼
You can orchestrate multi-agent swarms, run AST-based symbol search and blast-radius analysis, conduct boundary-first code reviews with scoring calibration, interrogate architecture before coding, select engineering personas like Ousterhout or Lamport, summarize sessions, and harvest ponytail debt comments into a ledger.
Who are these skills designed for?▼
Software engineers and technical leads managing agent-assisted development. They suit operators coordinating autonomous subagents, architects needing adversarial design interrogation, and developers wanting smoke-first TDD, minimal-diff iteration, and tracked technical debt across repositories.
How do I install and run these skills in a project?▼
Use the add-skills skill to opt a repository into suites from the shared library (interfaces, matt, specterops), or add-mcp-server to enable MCP servers (ast, browser, context7, context, akatsuki). Mimori and akatsuki run via native CLI and MCP stdio.
Are these skills tied to a specific model or harness?▼
No. Boundary-review is explicitly harness- and model-agnostic, operating via parallel subagents or sequential passes across any agent environment. Most skills set disable-model-invocation, meaning they trigger only on explicit user request rather than automatic model invocation.
What prerequisites or dependencies do these skills require?▼
Skills draw from a shared library of suites and MCP servers, enabled per-project through add-skills and add-mcp-server. Akatsuki requires access to the operator's Phoenix second brain for architecture orientation and work-log deposits; mimori needs no configuration for AST indexing.