[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CM4HJSTSP4FTM63TESAKJ1":3,"skill-same:01M7CM4HJSTSP4FTM63TESAKJ1":47},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":27,"commitSha":28,"skillDirectory":29,"forksCount":14,"repositoryHtmlUrl":30,"repositoryHomepageUrl":6,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":32,"install":33,"resources":46},"01M7CM4HJSTSP4FTM63TESAKJ1","mcp-builder",null,"MCP (Model Context Protocol) server and client engineering. Protocol design, schemas, transports, extensions, security, and migration.","haliltek\u002Fdev.tr","haliltek","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F31656259?v=4","01M7CKZCVHYRZ5CFNDBMFT5S7Q","Halil TEK","01M7CKZCWH3RX91D93NHE4DP2J",0,"AGPL-3.0","https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmcp-builder\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: mcp-builder\ndescription: MCP (Model Context Protocol) server and client engineering. Protocol design, schemas, transports, extensions, security, and migration.\nwhen_to_use: \"When building or reviewing MCP servers or clients, designing MCP tools\u002Fresources\u002Fprompts, migrating protocol versions, or validating MCP security and interoperability.\"\nallowed-tools: Read, Write, Edit, Glob, Grep\nversion: 1.0.0\n---\n\n# MCP Builder\n\n> Build interoperable MCP implementations against the stable **2026-07-28** specification. Treat extensions as negotiated, opt-in capabilities rather than core protocol guarantees.\n\n---\n\n## 1. Stable baseline and compatibility\n\n- The authoritative stable baseline is MCP **2026-07-28**.\n- Core requests are stateless and self-contained. Do not infer protocol version, client capabilities, identity, conversation, or task state from a transport connection or process lifetime.\n- Put cross-request state behind explicit identifiers supplied on every relevant request.\n- Negotiate the protocol version and capabilities; do not assume that a client or server implements every optional feature.\n- Preserve compatibility with earlier peers where the specification requires it, including accepting legacy resource-not-found error code `-32002` while emitting `-32602` for invalid parameters in the current protocol.\n\n### Core versus extensions\n\n| Layer | Status | Engineering rule |\n| --- | --- | --- |\n| Base protocol, versioning, message patterns | Stable core | Required for every implementation |\n| Resources, prompts, tools, elicitation | Optional core features | Advertise and check capabilities before use |\n| Tasks, Skills over MCP, MCP Apps | Extensions | Opt-in; require explicit support from both peers |\n| Draft or vendor extensions | Experimental\u002Fvendor-specific | Isolate behind adapters and feature flags |\n\nNever label an extension as stable merely because one SDK or host supports it. Check the extension's own official status and version.\n\n---\n\n## 2. Architecture\n\nMCP uses JSON-RPC 2.0 between hosts, clients, and servers.\n\n```text\nhost application\n  └─ MCP client \u002F runtime adapter\n       ├─ stdio transport\n       └─ Streamable HTTP transport\n            └─ MCP server\n                 ├─ resources\n                 ├─ prompts\n                 └─ tools\n```\n\n### Stateless request model\n\nEvery request must carry the protocol metadata required by the negotiated version. For 2026-07-28 this includes per-request protocol version and client capability metadata under reserved `_meta` keys.\n\nServers must implement `server\u002Fdiscover` for protocol-version and capability discovery. Clients may call it before normal requests or use it as a backward-compatibility probe for stdio peers.\n\nDo not:\n\n- use a stdio process as a conversation or session boundary;\n- depend on request ordering to establish capabilities or identity;\n- keep hidden mutable state without an explicit handle;\n- use self-reported client\u002Fserver information for authorization decisions.\n\nDo:\n\n- make task, thread, workspace, or subscription handles explicit;\n- validate every request independently;\n- tolerate interleaved requests from unrelated tasks;\n- document lifecycle, expiry, and cleanup for durable handles.\n\n---\n\n## 3. Transports and authentication\n\n| Transport | Typical use | Security baseline |\n| --- | --- | --- |\n| `stdio` | Local process integration | Read credentials from the environment; constrain process and filesystem access |\n| Streamable HTTP | Remote or shared service | Use the MCP HTTP authorization framework, validate origins where applicable, and enforce network boundaries |\n\nWebSocket is not a standard MCP transport unless defined by a separately negotiated extension or vendor adapter.\n\nTransport concerns must not leak into protocol semantics. Reconnection or process replacement must not destroy logical state that is represented by an explicit durable handle.\n\n---\n\n## 4. Schema design and validation\n\nMCP defaults to **JSON Schema 2020-12** when `$schema` is absent. Implementations must support that dialect and should document any additional dialects.\n\n### Tool input schemas\n\n- Use an object root for tool arguments.\n- Give every property a clear, action-oriented description.\n- Declare required fields explicitly.\n- Prefer narrow enums, bounds, formats, and `additionalProperties: false` where forward compatibility does not require open objects.\n- Validate both the schema and each invocation payload.\n\n### Tool output schemas\n\nUse structured content and `outputSchema` when consumers need typed, predictable results. Keep human-readable text concise and ensure structured output remains the source of truth for automation.\n\n### `$ref` and validator safety\n\n- Never dereference network `$ref` values automatically.\n- If remote dereferencing is explicitly enabled, use a host allowlist; reject loopback, link-local, and private addresses; apply timeouts and response-size limits; and log resolved URIs without secrets.\n- Reject schemas with unresolved external references instead of treating them as permissive.\n- Bound schema depth, total subschemas, and validation time to prevent denial-of-service through composition keywords or recursive references.\n\n---\n\n## 5. Tool, resource, and prompt design\n\n### Tools\n\n- Use clear action names and a single responsibility.\n- Mark read-only, destructive, idempotent, or open-world behavior with annotations when useful, but treat annotations as untrusted hints unless the server itself is trusted.\n- Require explicit user approval for consequential actions.\n- Return stable machine-readable errors without internal stack traces or secrets.\n\n### Resources\n\n- Use stable URIs and MIME types.\n- Separate discovery metadata from resource content.\n- Apply authorization per resource, not only per server connection.\n- Avoid silently forwarding resource data to another server or model without user consent.\n\n### Prompts\n\n- Treat prompts as user-visible templates, not hidden authority.\n- Keep tool permissions and policy enforcement outside prompt text.\n- Validate prompt arguments and clearly identify any data sources included in generated messages.\n\n---\n\n## 6. Security requirements\n\n1. **Consent and least privilege** — expose only required data and obtain explicit approval before tool execution or data sharing.\n2. **Untrusted metadata** — do not use tool descriptions, annotations, client info, server info, or model-produced arguments as authorization evidence.\n3. **Input and output validation** — validate schemas, arguments, structured results, URIs, and content types at trust boundaries.\n4. **Secret handling** — load credentials from environment or a secret manager; never place real keys in repository configuration, logs, traces, prompts, or error payloads.\n5. **Network controls** — defend remote schema\u002Fresource fetching against SSRF, redirect abuse, oversized responses, and slow responses.\n6. **Execution isolation** — sandbox subprocesses and filesystem access; use explicit path grants and reject archive or symlink escapes.\n7. **Auditability** — record security-relevant decisions, approvals, tool identity, and result status without storing sensitive payloads unnecessarily.\n\n---\n\n## 7. Extensions\n\nExtensions are negotiated independently from the core protocol.\n\n| Extension | Use | Guardrail |\n| --- | --- | --- |\n| Tasks | Long-running work, polling, durable handles, mid-flight input | Do not use the older experimental task API; negotiate the extension |\n| Skills over MCP | Discoverable structured agent instructions | Treat skill content as untrusted input and apply host policy |\n| MCP Apps | Interactive UI rendered in conversations | Constrain origins, content, data access, and action permissions |\n\nAn adapter must fail gracefully when an extension is absent. Never silently downgrade a security requirement to preserve feature parity.\n\n---\n\n## 8. Deprecated features and migration\n\nThe 2026-07-28 specification deprecates legacy roots, sampling, and logging shapes while keeping a compatibility window. Do not remove compatibility abruptly, but avoid designing new architecture around deprecated forms.\n\n### Migration checklist from 2025-era implementations\n\n- [ ] Replace connection-scoped sessions and the `initialize` handshake with stateless per-request metadata.\n- [ ] Implement or consume `server\u002Fdiscover` for version and capability selection.\n- [ ] Move durable state behind explicit task, thread, workspace, or application handles.\n- [ ] Send and validate negotiated protocol version and client capabilities on every request.\n- [ ] Upgrade schema validation to JSON Schema 2020-12 by default.\n- [ ] Disable network `$ref` resolution by default and add validation resource limits.\n- [ ] Move experimental Tasks usage to the negotiated Tasks extension.\n- [ ] Accept legacy `-32002` resource-not-found responses, but emit current error semantics.\n- [ ] Treat a missing `resultType` from older servers as `\"complete\"`; require it for current-version results.\n- [ ] Audit roots, sampling, and logging integrations for their documented replacements and deprecation timeline.\n- [ ] Add interoperability tests against at least one older peer and one 2026-07-28 peer.\n\n---\n\n## 9. Testing matrix\n\n| Test | Expected evidence |\n| --- | --- |\n| Contract | Invalid JSON-RPC, missing metadata, unsupported version\u002Fcapability, schema dialect handling |\n| Statelessness | Interleaved requests, reconnects, process reuse, explicit-handle recovery |\n| Schema security | External `$ref`, recursive\u002Fcomposed schemas, depth\u002Fcount\u002Ftime limits |\n| Authorization | Denied resource\u002Ftool access, approval gates, credential isolation |\n| Extension negotiation | Extension absent, version mismatch, feature enabled, graceful fallback |\n| Compatibility | Current peer plus supported older protocol revision |\n| Failure handling | Timeouts, cancellation, partial results, structured errors, no secret leakage |\n\n---\n\n## 10. Review checklist\n\n- [ ] Stable protocol version and extension status are documented.\n- [ ] Requests are stateless; cross-request state uses explicit handles.\n- [ ] Capabilities are negotiated and checked before optional operations.\n- [ ] JSON Schema 2020-12 is supported and validator resource use is bounded.\n- [ ] Network `$ref` resolution is disabled by default.\n- [ ] Tool annotations are not trusted as authorization.\n- [ ] User approval, least privilege, secret handling, and execution isolation are enforced outside prompts.\n- [ ] Deprecated features have a compatibility and migration plan.\n- [ ] Contract, security, extension, and cross-version tests pass.\n","# MCP Builder\n\n> Build interoperable MCP implementations against the stable **2026-07-28** specification. Treat extensions as negotiated, opt-in capabilities rather than core protocol guarantees.\n\n---\n\n## 1. Stable baseline and compatibility\n\n- The authoritative stable baseline is MCP **2026-07-28**.\n- Core requests are stateless and self-contained. Do not infer protocol version, client capabilities, identity, conversation, or task state from a transport connection or process lifetime.\n- Put cross-request state behind explicit identifiers supplied on every relevant request.\n- Negotiate the protocol version and capabilities; do not assume that a client or server implements every optional feature.\n- Preserve compatibility with earlier peers where the specification requires it, including accepting legacy resource-not-found error code `-32002` while emitting `-32602` for invalid parameters in the current protocol.\n\n### Core versus extensions\n\n| Layer | Status | Engineering rule |\n| --- | --- | --- |\n| Base protocol, versioning, message patterns | Stable core | Required for every implementation |\n| Resources, prompts, tools, elicitation | Optional core features | Advertise and check capabilities before use |\n| Tasks, Skills over MCP, MCP Apps | Extensions | Opt-in; require explicit support from both peers |\n| Draft or vendor extensions | Experimental\u002Fvendor-specific | Isolate behind adapters and feature flags |\n\nNever label an extension as stable merely because one SDK or host supports it. Check the extension's own official status and version.\n\n---\n\n## 2. Architecture\n\nMCP uses JSON-RPC 2.0 between hosts, clients, and servers.\n\n```text\nhost application\n  └─ MCP client \u002F runtime adapter\n       ├─ stdio transport\n       └─ Streamable HTTP transport\n            └─ MCP server\n                 ├─ resources\n                 ├─ prompts\n                 └─ tools\n```\n\n### Stateless request model\n\nEvery request must carry the protocol metadata required by the negotiated version. For 2026-07-28 this includes per-request protocol version and client capability metadata under reserved `_meta` keys.\n\nServers must implement `server\u002Fdiscover` for protocol-version and capability discovery. Clients may call it before normal requests or use it as a backward-compatibility probe for stdio peers.\n\nDo not:\n\n- use a stdio process as a conversation or session boundary;\n- depend on request ordering to establish capabilities or identity;\n- keep hidden mutable state without an explicit handle;\n- use self-reported client\u002Fserver information for authorization decisions.\n\nDo:\n\n- make task, thread, workspace, or subscription handles explicit;\n- validate every request independently;\n- tolerate interleaved requests from unrelated tasks;\n- document lifecycle, expiry, and cleanup for durable handles.\n\n---\n\n## 3. Transports and authentication\n\n| Transport | Typical use | Security baseline |\n| --- | --- | --- |\n| `stdio` | Local process integration | Read credentials from the environment; constrain process and filesystem access |\n| Streamable HTTP | Remote or shared service | Use the MCP HTTP authorization framework, validate origins where applicable, and enforce network boundaries |\n\nWebSocket is not a standard MCP transport unless defined by a separately negotiated extension or vendor adapter.\n\nTransport concerns must not leak into protocol semantics. Reconnection or process replacement must not destroy logical state that is represented by an explicit durable handle.\n\n---\n\n## 4. Schema design and validation\n\nMCP defaults to **JSON Schema 2020-12** when `$schema` is absent. Implementations must support that dialect and should document any additional dialects.\n\n### Tool input schemas\n\n- Use an object root for tool arguments.\n- Give every property a clear, action-oriented description.\n- Declare required fields explicitly.\n- Prefer narrow enums, bounds, formats, and `additionalProperties: false` where forward compatibility does not require open objects.\n- Validate both the schema and each invocation payload.\n\n### Tool output schemas\n\nUse structured content and `outputSchema` when consumers need typed, predictable results. Keep human-readable text concise and ensure structured output remains the source of truth for automation.\n\n### `$ref` and validator safety\n\n- Never dereference network `$ref` values automatically.\n- If remote dereferencing is explicitly enabled, use a host allowlist; reject loopback, link-local, and private addresses; apply timeouts and response-size limits; and log resolved URIs without secrets.\n- Reject schemas with unresolved external references instead of treating them as permissive.\n- Bound schema depth, total subschemas, and validation time to prevent denial-of-service through composition keywords or recursive references.\n\n---\n\n## 5. Tool, resource, and prompt design\n\n### Tools\n\n- Use clear action names and a single responsibility.\n- Mark read-only, destructive, idempotent, or open-world behavior with annotations when useful, but treat annotations as untrusted hints unless the server itself is trusted.\n- Require explicit user approval for consequential actions.\n- Return stable machine-readable errors without internal stack traces or secrets.\n\n### Resources\n\n- Use stable URIs and MIME types.\n- Separate discovery metadata from resource content.\n- Apply authorization per resource, not only per server connection.\n- Avoid silently forwarding resource data to another server or model without user consent.\n\n### Prompts\n\n- Treat prompts as user-visible templates, not hidden authority.\n- Keep tool permissions and policy enforcement outside prompt text.\n- Validate prompt arguments and clearly identify any data sources included in generated messages.\n\n---\n\n## 6. Security requirements\n\n1. **Consent and least privilege** — expose only required data and obtain explicit approval before tool execution or data sharing.\n2. **Untrusted metadata** — do not use tool descriptions, annotations, client info, server info, or model-produced arguments as authorization evidence.\n3. **Input and output validation** — validate schemas, arguments, structured results, URIs, and content types at trust boundaries.\n4. **Secret handling** — load credentials from environment or a secret manager; never place real keys in repository configuration, logs, traces, prompts, or error payloads.\n5. **Network controls** — defend remote schema\u002Fresource fetching against SSRF, redirect abuse, oversized responses, and slow responses.\n6. **Execution isolation** — sandbox subprocesses and filesystem access; use explicit path grants and reject archive or symlink escapes.\n7. **Auditability** — record security-relevant decisions, approvals, tool identity, and result status without storing sensitive payloads unnecessarily.\n\n---\n\n## 7. Extensions\n\nExtensions are negotiated independently from the core protocol.\n\n| Extension | Use | Guardrail |\n| --- | --- | --- |\n| Tasks | Long-running work, polling, durable handles, mid-flight input | Do not use the older experimental task API; negotiate the extension |\n| Skills over MCP | Discoverable structured agent instructions | Treat skill content as untrusted input and apply host policy |\n| MCP Apps | Interactive UI rendered in conversations | Constrain origins, content, data access, and action permissions |\n\nAn adapter must fail gracefully when an extension is absent. Never silently downgrade a security requirement to preserve feature parity.\n\n---\n\n## 8. Deprecated features and migration\n\nThe 2026-07-28 specification deprecates legacy roots, sampling, and logging shapes while keeping a compatibility window. Do not remove compatibility abruptly, but avoid designing new architecture around deprecated forms.\n\n### Migration checklist from 2025-era implementations\n\n- [ ] Replace connection-scoped sessions and the `initialize` handshake with stateless per-request metadata.\n- [ ] Implement or consume `server\u002Fdiscover` for version and capability selection.\n- [ ] Move durable state behind explicit task, thread, workspace, or application handles.\n- [ ] Send and validate negotiated protocol version and client capabilities on every request.\n- [ ] Upgrade schema validation to JSON Schema 2020-12 by default.\n- [ ] Disable network `$ref` resolution by default and add validation resource limits.\n- [ ] Move experimental Tasks usage to the negotiated Tasks extension.\n- [ ] Accept legacy `-32002` resource-not-found responses, but emit current error semantics.\n- [ ] Treat a missing `resultType` from older servers as `\"complete\"`; require it for current-version results.\n- [ ] Audit roots, sampling, and logging integrations for their documented replacements and deprecation timeline.\n- [ ] Add interoperability tests against at least one older peer and one 2026-07-28 peer.\n\n---\n\n## 9. Testing matrix\n\n| Test | Expected evidence |\n| --- | --- |\n| Contract | Invalid JSON-RPC, missing metadata, unsupported version\u002Fcapability, schema dialect handling |\n| Statelessness | Interleaved requests, reconnects, process reuse, explicit-handle recovery |\n| Schema security | External `$ref`, recursive\u002Fcomposed schemas, depth\u002Fcount\u002Ftime limits |\n| Authorization | Denied resource\u002Ftool access, approval gates, credential isolation |\n| Extension negotiation | Extension absent, version mismatch, feature enabled, graceful fallback |\n| Compatibility | Current peer plus supported older protocol revision |\n| Failure handling | Timeouts, cancellation, partial results, structured errors, no secret leakage |\n\n---\n\n## 10. Review checklist\n\n- [ ] Stable protocol version and extension status are documented.\n- [ ] Requests are stateless; cross-request state uses explicit handles.\n- [ ] Capabilities are negotiated and checked before optional operations.\n- [ ] JSON Schema 2020-12 is supported and validator resource use is bounded.\n- [ ] Network `$ref` resolution is disabled by default.\n- [ ] Tool annotations are not trusted as authorization.\n- [ ] User approval, least privilege, secret handling, and execution isolation are enforced outside prompts.\n- [ ] Deprecated features have a compatibility and migration plan.\n- [ ] Contract, security, extension, and cross-version tests pass.\n",{"name":5,"version":24,"description":7,"when_to_use":25,"allowed-tools":26},"1.0.0","When building or reviewing MCP servers or clients, designing MCP tools\u002Fresources\u002Fprompts, migrating protocol versions, or validating MCP security and interoperability.","Read, Write, Edit, Glob, Grep","main","35fc33da79b31d7f8dadfc704083baff1b5672a0",".agents\u002Fskills\u002Fmcp-builder","https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr","https:\u002F\u002Fgithub.com\u002Fhaliltek","2026-11-13 01:38:48",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add haliltek\u002Fdev.tr --skill mcp-builder",{"tool":39,"command":40},"bunx","bunx skills add haliltek\u002Fdev.tr --skill mcp-builder",{"tool":42,"command":43},"pnpm","pnpm dlx skills add haliltek\u002Fdev.tr --skill mcp-builder","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fmcp-builder\u002FSKILL.md",true,[],[48,55,62,69,76,83],{"skillNo":49,"name":50,"nameZh":6,"description":51,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":52,"categories":53,"keywords":54,"publishedAt":19},"01M7CM8JAC0X6HVS8VCHH51KPZ","respectfully","Wraps any prompt with a reminder to respect CLAUDE.md files","https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Frespectfully\u002FSKILL.md",[],[],{"skillNo":56,"name":57,"nameZh":6,"description":58,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":59,"categories":60,"keywords":61,"publishedAt":19},"01M7CM8FPHK4J5EP607Y77XHA4","recruiter","Think like a skeptical technical recruiter when working on the daily.dev Recruiter platform. Use when reviewing UI\u002FUX, features, or code in webapp\u002Fpages\u002Frecruiter.","https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Frecruiter\u002FSKILL.md",[],[],{"skillNo":63,"name":64,"nameZh":6,"description":65,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":66,"categories":67,"keywords":68,"publishedAt":19},"01M7CM8CB9XR2W46XJ3T2X0245","changelog","Generate a daily.dev changelog post for a new feature or update. Use when a team member needs to write a changelog announcement.","https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fchangelog\u002FSKILL.md",[],[],{"skillNo":70,"name":71,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":19},"01M7CM88HKHZP4S2H9GYKWCBPB","webapp-testing","Web application testing principles. E2E, Playwright, deep audit strategies.","https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fwebapp-testing\u002FSKILL.md",[],[],{"skillNo":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":19},"01M7CM831R2XYCHFNRGXD6NYS1","web-design-guidelines","Review UI code for Web Interface Guidelines compliance. Use when asked to \"review my UI\", \"check accessibility\", \"audit design\", \"review UX\", or \"check my site against best practices\".","https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fweb-design-guidelines\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":19},"01M7CM7XTJ13ZQFZ9Y5SRYF8CP","vulnerability-scanner","Advanced vulnerability analysis principles. OWASP 2025, Supply Chain Security, attack surface mapping, risk prioritization.","https:\u002F\u002Fgithub.com\u002Fhaliltek\u002Fdev.tr\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fvulnerability-scanner\u002FSKILL.md",[],[]]