Connector compatibility
Some connectors work today. Some need account setup or admin approval. Others are listed so you know they are planned, not live. Each card says what HeapFile can read, where the data travels, and what it will not do.
Provider names identify systems HeapFile can work with or is evaluating for compatibility. HeapFile is not affiliated with, endorsed by, or sponsored by these providers unless explicitly stated. Availability depends on the HeapFile release, provider terms, account plan, organization policy, authorization, and configuration.
HeapFile is local-first, not local-only. Local sources and eligible local model routes can remain on the device. A cloud connector contacts its provider only through a route the user or administrator authorizes. The product identifies the permission level and data route before connection.
HeapFile-owned connectors are separate from ChatGPT, Codex, Claude, browser tabs, local CLI state, and local checkouts. Chat-client connectors do not authorize HeapFile. A listed connector is not an authorized connector.
Default view
Choose explicit safe workspace roots.
Uses: Selected local files and documents.
Does not: Crawl the whole home folder by default.
Use repositories already present on this device.
Uses: Local repository history, diffs, branches, and file evidence.
Does not: Authorize hosted Git providers.
Register HeapFile MCP with a compatible local AI client.
Uses: Work Memory, Handoff Packs, and local HeapFile tools.
Does not: Force every client to call HeapFile automatically.
Install and run an eligible local model route.
Uses: Local model route for bounded local reasoning.
Does not: Imply all model providers or mobile local models are supported.
Install the read-only HeapFile GitHub App on selected repositories.
Uses: Repository, issue, pull request, commit, release, workflow, and check evidence where authorized.
Does not: Request write, administration, webhook, secret, or deployment permissions.
Needs setup or more proof
Configure a tenant and sign in through HeapFile's Atlassian browser session.
Uses: Issue search, issue fetch, and ticket evidence.
Does not: Update tickets, add comments, or claim marketplace-app availability.
Use the same authorized Atlassian browser-session path.
Uses: Space and page search plus page evidence.
Does not: Create, update, delete, or publish pages.
Configure an organization and verify a browser-session read path.
Uses: Projects, repositories, pull requests, work items, pipelines, and builds.
Does not: Expose release approvals or mutations through the evidence connector.
Configure a runtime-only read token or OAuth path.
Uses: Projects, issues, merge requests, and commits.
Does not: Claim hosted customer OAuth rollout yet.
Configure read-only Bitbucket Cloud OAuth or access-token authorization.
Uses: Workspaces, repositories, pull requests, issues, and commits.
Does not: Claim broad customer workspace search yet.
Breadth without overclaiming
Requires workspace authorization, channel allowlists, and read-only scopes.
Uses: Planned message, decision, and follow-up evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a HeapFile-owned Google authorization path with scoped read access and explicit file or drive boundaries.
Uses: Planned document, spreadsheet, slide, and file evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a HeapFile-owned Google mail read path and account or workspace approval.
Uses: Planned thread evidence for obligations and decisions.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires approved app or bot authorization plus explicit server and channel allowlists.
Uses: Planned community and support-channel evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires an approved Notion integration path with selected page or database access.
Uses: Planned page and database context for product specs and operating notes.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a HeapFile-owned Linear authorization path with issue and cycle read scopes.
Uses: Planned issue, project, and cycle evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires workspace selection, project allowlists, and task read scopes.
Uses: Planned task ownership, due date, and project status evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires workspace or board allowlists and read-only board and card scopes.
Uses: Planned board, list, and card evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a customer-approved connected app, object allowlists, and read scopes.
Uses: Planned account, opportunity, and case evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires an approved private app or OAuth path with CRM object scopes.
Uses: Planned company, contact, deal, and ticket evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a Zendesk subdomain allowlist and read-only ticket or search scopes.
Uses: Planned ticket and help-center evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires approved conversation and contact read scopes with workspace boundaries.
Uses: Planned support and success conversation evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a Microsoft Graph read path or approved organization bridge.
Uses: Planned mailbox evidence for decisions and follow-ups.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires an approved read-only calendar source.
Uses: Planned event context and scheduling evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a scoped Microsoft read path and selected folders or files.
Uses: Planned selected document evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires Microsoft Graph or an approved organization bridge with site and file boundaries.
Uses: Planned organization document evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires approved Microsoft 365 access and clear channel or chat boundaries.
Uses: Planned meeting, channel, and follow-up context.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a future Google Cloud source-control read path with project and repository allowlists.
Uses: Coverage planning for legacy or Google Cloud-centered source repositories.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires future explicit instance URL allowlists, TLS checks, and read-only tokens.
Uses: Coverage planning for self-hosted source control.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires future customer-approved Sourcegraph endpoints, search tokens, and query bounds.
Uses: Research coverage for indexed code-search evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a HeapFile-owned IAM Identity Center or scoped role path with account allowlists.
Uses: Planned account, deployment, cost, and audit evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires an Azure app or workload identity with tenant and subscription allowlists.
Uses: Planned subscription, resource, deployment, cost, and activity-log evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a Google Cloud authorization path with project allowlists and read-only asset, billing, or logging scopes.
Uses: Planned project, asset, billing, and audit evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires explicit kubeconfig or context selection, read-only RBAC, and namespace allowlists.
Uses: Planned cluster, namespace, workload, event, and status evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires explicit read-only connection profiles, table allowlists, and query bounds.
Uses: Planned schema, table, and bounded query evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires read-only authorization, role and warehouse allowlists, and bounded evidence queries.
Uses: Planned warehouse, schema, and analytics context.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires read-only API and application keys, site selection, service allowlists, and query bounds.
Uses: Planned service, metric, log, monitor, and incident evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires explicit endpoints, TLS policy, read-only service accounts, datasource allowlists, and query limits.
Uses: Planned metrics, dashboards, and incident context.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires organization and project allowlists plus read-only issue, release, and event scopes.
Uses: Planned issue, release, trace, and event evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires account or zone allowlists and a read-only API path.
Uses: Planned zone, DNS, Workers, Pages, WAF, cache, and analytics evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires explicit local or remote backend profiles, workspace allowlists, and state redaction.
Uses: Planned workspace, state, drift, and plan metadata.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a provider-owned usage or project metadata path approved for HeapFile.
Uses: Planned project, usage, and cost-evidence metadata.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires an approved organization or workspace read path for metadata and usage evidence.
Uses: Planned usage and workspace metadata.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires a Google AI project or usage read path approved for HeapFile.
Uses: Planned project and usage evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires scoped AWS authorization with Bedrock and cost evidence boundaries.
Uses: Planned model, agent, and cost evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires an approved read path for account, model, or usage metadata.
Uses: Planned model-routing and usage evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires explicit organization or user allowlists and a read-only authorization path.
Uses: Planned model, dataset, and Spaces metadata.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires account or team allowlists and read-only model or prediction metadata scopes.
Uses: Planned hosted model experiment evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires future YouTube Data API, OAuth, or approved local export routes with channel allowlists and quota bounds.
Uses: Planned channel, video, transcript, and creator workflow evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires future Twitch authorization, channel allowlists, and bounded VOD or clip retrieval.
Uses: Planned channel, stream, VOD, and clip evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires platform-policy and product practicality review plus approved APIs or exports.
Uses: Research coverage for short-form creator evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires approved Meta or Instagram routes, account allowlists, and policy review.
Uses: Research coverage for creator and short-form media evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Requires approved Meta or Facebook routes, page allowlists, and policy review.
Uses: Research coverage for page and video evidence.
Does not: Make live API calls, discover credentials, scrape, write, or reuse chat-client connectors.
Mobile status
Phones and tablets are for capture, selected recall, approval review, and handoff. Public store availability depends on store review and release status, so mobile is described separately from provider connectors.