# Send GraphQL to your agent Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually. ## Fast path - Download the package from Yavira. - Extract it into a folder your agent can access. - Paste one of the prompts below and point your agent at the extracted folder. ## Suggested prompts ### New install ```text I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete. ``` ### Upgrade existing ```text I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run. ``` ## Machine-readable fields ```json { "schemaVersion": "1.0", "item": { "slug": "graphql", "name": "GraphQL", "source": "tencent", "type": "skill", "category": "AI 智能", "sourceUrl": "https://clawhub.ai/ivangdavila/graphql", "canonicalUrl": "https://clawhub.ai/ivangdavila/graphql", "targetPlatform": "OpenClaw" }, "install": { "downloadUrl": "/downloads/graphql", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=graphql", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "packageFormat": "ZIP package", "primaryDoc": "SKILL.md", "includedAssets": [ "SKILL.md", "client.md", "performance.md", "schema.md", "security.md" ], "downloadMode": "redirect", "sourceHealth": { "source": "tencent", "slug": "graphql", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-05-01T18:54:06.379Z", "expiresAt": "2026-05-08T18:54:06.379Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=graphql", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=graphql", "contentDisposition": "attachment; filename=\"graphql-1.0.1.zip\"", "redirectLocation": null, "bodySnippet": null, "slug": "graphql" }, "scope": "item", "summary": "Item download looks usable.", "detail": "Yavira can redirect you to the upstream package for this item.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/graphql" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] } }, "links": { "detailUrl": "https://openagent3.xyz/skills/graphql", "downloadUrl": "https://openagent3.xyz/downloads/graphql", "agentUrl": "https://openagent3.xyz/skills/graphql/agent", "manifestUrl": "https://openagent3.xyz/skills/graphql/agent.json", "briefUrl": "https://openagent3.xyz/skills/graphql/agent.md" } } ``` ## Documentation ### Quick Reference TopicFileSchema design patternsschema.mdSecurity and limitssecurity.mdPerformance optimizationperformance.mdClient-side patternsclient.md ### N+1 Problem (Critical) Each resolver runs independently—fetching user for each of 100 posts = 100 queries DataLoader required: batches requests within single tick—100 posts = 1 user query DataLoader per-request: create new instance per request—prevents cross-request caching Even with DataLoader, watch for nested N+1—posts → comments → authors chains ### Schema Design Fields nullable by default—make non-null explicit: name: String! Input types separate from output—CreateUserInput vs User; allows different validation Connections for pagination: users(first: 10, after: "cursor") returns edges + pageInfo Avoid deeply nested types—flatten where possible; 5+ levels = resolver complexity ### Pagination Cursor-based (Relay style): first/after, last/before—stable across insertions Offset-based: limit/offset—simpler but skips or duplicates on concurrent writes Return pageInfo { hasNextPage, endCursor }—client knows when to stop totalCount expensive on large datasets—make optional or estimate ### Security Traps Query depth limiting—prevent { user { friends { friends { friends... } } } } Query complexity scoring—count fields, multiply by list sizes; reject above threshold Disable introspection in production—or protect with auth; schema is attack surface Timeout per query—malicious queries can be slow without being deep Rate limit by complexity, not just requests—one complex query = many simple ones ### Error Handling Partial success normal—query returns data AND errors; check both Errors array with path—shows which field failed: "path": ["user", "posts", 0] Error extensions for codes—"extensions": {"code": "FORBIDDEN"}; don't parse message Throw in resolver = null + error—parent nullable = partial data; parent non-null = error propagates up ### Resolver Patterns Return object with ID, let sub-resolvers fetch details—avoids over-fetching at top level __resolveType for unions/interfaces—required to determine concrete type Context for auth, DataLoaders, DB connection—pass through resolver chain Field-level auth in resolvers—check permissions per field, not just per query ### Mutations Return modified object—client updates cache without re-fetch Input validation before DB—return user-friendly error, not DB constraint violation Idempotency for critical mutations—accept client-generated ID or idempotency key One mutation per operation typically—batch mutations exist but complicate error handling ### Performance Persisted queries: hash → query mapping—smaller payloads, prevents arbitrary queries @defer for slow fields—returns fast fields first, streams slow ones (if supported) Fragment colocation: components define data needs—reduces over-fetching Query allowlisting: only registered queries in production—blocks exploratory attacks ### Subscriptions WebSocket-based—graphql-ws protocol; separate from HTTP Scaling: pub/sub needed—Redis or similar for multi-server broadcast Filter at subscription level—don't push everything and filter client-side Unsubscribe on disconnect—clean up resources; connection tracking required ### Client-Side Normalized cache (Apollo, Relay)—deduplicate by ID; updates propagate automatically Optimistic UI: predict mutation result—rollback if server differs Error policies: decide per-query—ignore errors, return partial, or treat as failure Fragment reuse—define once, use in multiple queries; keeps fields in sync ### Common Mistakes No DataLoader—N+1 kills performance; one query becomes hundreds Exposing internal errors—stack traces leak implementation details No query limits—attackers craft expensive queries; DoS with single request Over-fetching in resolvers—fetching full object when query only needs ID + name Treating like REST—GraphQL is a graph; design for traversal, not resources ## Trust - Source: tencent - Verification: Indexed source record - Publisher: ivangdavila - Version: 1.0.1 ## Source health - Status: healthy - Item download looks usable. - Yavira can redirect you to the upstream package for this item. - Health scope: item - Reason: direct_download_ok - Checked at: 2026-05-01T18:54:06.379Z - Expires at: 2026-05-08T18:54:06.379Z - Recommended action: Download for OpenClaw ## Links - [Detail page](https://openagent3.xyz/skills/graphql) - [Send to Agent page](https://openagent3.xyz/skills/graphql/agent) - [JSON manifest](https://openagent3.xyz/skills/graphql/agent.json) - [Markdown brief](https://openagent3.xyz/skills/graphql/agent.md) - [Download page](https://openagent3.xyz/downloads/graphql)