chore(toolkit-docs): restore lost summaries for 6 toolkits

toolkit-docs-generator/data/toolkits/daytona.json

@@ -4192,5 +4192,6 @@
   "documentationChunks": [],
   "customImports": [],
   "subPages": [],
-  "generatedAt": "2026-04-02T11:25:02.814Z"
-}
+  "generatedAt": "2026-04-20T17:27:26.000Z",
+  "summary": "Arcade's Daytona toolkit lets LLMs provision and manage isolated sandboxes, run code and commands, operate persistent sessions, manage snapshots, and perform Git and file operations inside the sandbox. It streamlines dev-loop automation including port previews and SSH access.\n\n**Capabilities**\n- Full sandbox lifecycle: create (from snapshot or image), start, stop, archive, delete, resize, label, list regions, and configure auto-stop / auto-archive / auto-delete intervals.\n- File system and content tools: read, write, move, delete, list, get info, find by glob, search content, replace across files, and create folders.\n- Git operations inside the sandbox: clone, status, add, commit, push, pull, log, checkout, and branch create/list/delete.\n- Execute code or shell commands directly, or drive persistent sessions with input, logs, and listings for long-running processes.\n- Access and sharing: snapshot create/list/get, port listing and preview URLs (standard or signed), and time-boxed SSH access credentials.\n\n**OAuth**\nUses GitHub OAuth for user identification. See the [Arcade GitHub auth provider docs](https://docs.arcade.dev/en/references/auth-providers/github) for configuration.\n\n**Secrets**\n- `DAYTONA_API_KEY` — authenticates Arcade to the Daytona API. Configure in the [Arcade Dashboard](https://api.arcade.dev/dashboard/auth/secrets) per the [Arcade secret setup guide](https://docs.arcade.dev/en/guides/create-tools/tool-basics/create-tool-secrets). SSH tokens returned by `CreateSshAccess` are sensitive and should not be logged."
+}

toolkit-docs-generator/data/toolkits/github.json

@@ -5766,5 +5766,6 @@
     "import { Callout, Tabs } from \"nextra/components\";"
   ],
   "subPages": [],
-  "generatedAt": "2026-04-17T11:27:15.538Z"
-}
+  "generatedAt": "2026-04-17T11:27:15.538Z",
+  "summary": "Arcade's GitHub toolkit provides authenticated access to GitHub resources, enabling automated repository, issue, pull request, review, and Projects V2 workflows. It exposes high-level actions that handle fuzzy matching, diff-aware edits, and user-centric queries without forcing agents to juggle raw REST calls.\n\n**Capabilities**\n- Manage repository and file operations: create branches, read files with line ranges, create or overwrite files, and replace line blocks safely.\n- Drive the pull request lifecycle: create, update, list, merge, check merge readiness, manage reviewers, submit reviews, and reply to or resolve review threads.\n- Administer issues and labels with fuzzy-matched add/remove, comments, assignment, and issue-to-project linking.\n- Query Projects V2 fields and items, list stargazers, collaborators, labels, activities, and authenticated-user views (open items, recent activity, review workload).\n\n**OAuth**\nRequires GitHub OAuth. See the [Arcade GitHub auth provider docs](https://docs.arcade.dev/en/references/auth-providers/github) for configuration.\n\n**Secrets**\n- `GITHUB_SERVER_URL` — override for GitHub Enterprise endpoints. Configure in the [Arcade Dashboard](https://api.arcade.dev/dashboard/auth/secrets) per the [Arcade secret setup guide](https://docs.arcade.dev/en/guides/create-tools/tool-basics/create-tool-secrets)."
+}

toolkit-docs-generator/data/toolkits/googledocs.json

@@ -1175,5 +1175,6 @@
     "import ScopePicker from \"@/app/_components/scope-picker\";"
   ],
   "subPages": [],
-  "generatedAt": "2026-04-10T11:26:08.782Z"
-}
+  "generatedAt": "2026-04-20T17:27:26.000Z",
+  "summary": "Arcade's Google Docs toolkit lets LLM-driven apps create, read, edit, comment on, and search Google Docs while accessing document metadata and authenticated user profile information. It understands multi-tab documents and returns structured DocMD output for context-aware editing.\n\n**Capabilities**\n- Create documents from blank templates or from plain text / Markdown, and append or edit body content via batchUpdate.\n- Retrieve structured DocMD with tab annotations, fetch document metadata with hierarchical tab character counts, or search and retrieve documents with body content.\n- Search Drive for documents (metadata-only or with content), list and add comments, and generate a Google File Picker URL to recover from missing-file or permission errors.\n- Expose authenticated user profile and Docs environment info through WhoAmI.\n\n**OAuth**\nRequires Google OAuth. See the [Arcade Google auth provider docs](https://docs.arcade.dev/en/references/auth-providers/google) for configuration.\n\n**Secrets**\n- `OPENAI_API_KEY` — used for LLM-assisted edit parsing. Configure in the [Arcade Dashboard](https://api.arcade.dev/dashboard/auth/secrets) per the [Arcade secret setup guide](https://docs.arcade.dev/en/guides/create-tools/tool-basics/create-tool-secrets)."
+}

toolkit-docs-generator/data/toolkits/jira.json

@@ -4575,5 +4575,6 @@
       "relativePath": "environment-variables/page.mdx"
     }
   ],
-  "generatedAt": "2026-04-17T11:27:15.538Z"
-}
+  "generatedAt": "2026-04-17T11:27:15.538Z",
+  "summary": "Arcade's Jira toolkit enables LLM-driven interactions with Atlassian Jira, letting agents create and update issues, move work through transitions, manage sprints and boards, handle attachments and comments, and resolve projects, users, and priorities. It is designed for programmatic issue lifecycle and Agile workflows with aggressive single-call batching.\n\n**Capabilities**\n- Manage issues end-to-end: create, update, transition, comment, add/remove labels, and attach or download files.\n- Run Scrum/Kanban workflows: list boards and sprints with date filters, fetch sprint and backlog issues, add issues to sprints, and move issues back to the backlog.\n- Search and browse flexibly: parameterized issue search (preferred) or raw JQL, plus listings for projects, users, issue types, priorities, and priority schemes.\n- Resolve references by name, key, or email so agents avoid unnecessary lookup round-trips, and establish user context with WhoAmI against available Atlassian clouds.\n\n**OAuth**\nRequires Atlassian OAuth. See the [Arcade Atlassian auth provider docs](https://docs.arcade.dev/en/references/auth-providers/atlassian) for configuration."
+}

toolkit-docs-generator/data/toolkits/linear.json

@@ -4066,5 +4066,6 @@
   ],
   "customImports": [],
   "subPages": [],
-  "generatedAt": "2026-04-18T11:20:23.013Z"
-}
+  "generatedAt": "2026-04-18T11:20:23.013Z",
+  "summary": "Arcade's Linear toolkit lets LLMs interact with Linear workspaces, enabling programmatic creation, update, archival, and search across issues, projects, initiatives, cycles, and comments. It validates entity references and exposes workspace metadata so agents can plan work without manual lookups.\n\n**Capabilities**\n- Full issue lifecycle: create, update, archive, transition workflow state, relate issues, and link GitHub PRs, commits, or issues.\n- Project and initiative management: create, update, archive, post status updates, link projects to initiatives, and paginate long descriptions.\n- Collaboration: add, list, update, and reply to comments on issues and project documents; manage per-issue notification subscriptions.\n- Discovery and context: list and get cycles, teams, labels, workflow states, notifications, and recent authenticated-user activity; resolve the current user via WhoAmI.\n\n**OAuth**\nRequires Linear OAuth. See the [Arcade Linear auth provider docs](https://docs.arcade.dev/en/references/auth-providers/linear) for configuration."
+}

toolkit-docs-generator/data/toolkits/salesforce.json

@@ -2381,5 +2381,6 @@
   "documentationChunks": [],
   "customImports": [],
   "subPages": [],
-  "generatedAt": "2026-04-02T11:25:02.802Z"
-}
+  "generatedAt": "2026-04-20T17:27:26.000Z",
+  "summary": "Arcade's Salesforce toolkit lets LLMs interact with Salesforce orgs to create, update, search, and convert CRM records, log activities, and fetch enriched relational data. It validates against org-configured picklists and returns contextual warnings and IDs to guide multi-step agent workflows.\n\n**Capabilities**\n- CRUD and lifecycle for leads, contacts, opportunities, and tasks, including the canonical lead-to-deal conversion with Contact/Account/Opportunity creation.\n- Rich account and opportunity retrieval with related contacts, leads, notes, calls, tasks, emails, events, contact roles, and line items.\n- Filtered search across accounts, contacts, leads, and opportunities with owner-scoped (me/all) and paginated results.\n- Activity logging and task management: log completed calls, create follow-up tasks, and list or update assigned tasks with overdue filters.\n- Org-aware validation: picklist checks, conversion guards, and WhoAmI for authenticated user and org context.\n\n**OAuth**\nRequires Salesforce OAuth. See the [Arcade Salesforce auth provider docs](https://docs.arcade.dev/en/references/auth-providers/salesforce) for configuration.\n\n**Secrets**\n- `SALESFORCE_ORG_SUBDOMAIN` — target org subdomain.\n- `SALESFORCE_MAX_CONCURRENT_REQUESTS` — concurrency limit for batched related-record fetches.\n\nConfigure both in the [Arcade Dashboard](https://api.arcade.dev/dashboard/auth/secrets) per the [Arcade secret setup guide](https://docs.arcade.dev/en/guides/create-tools/tool-basics/create-tool-secrets)."
+}