Add a third-party MCP server (Linear, Notion, GitHub, ...) to the PostHog MCP store catalog. Use when asked to "add X to the MCP store", expand the MCP server marketplace, or fix a broken catalog entry. Covers finding the vendor's remote MCP endpoint, probing it (handshake, OAuth discovery, DCR), authoring the catalog entry in products/mcp_store/backend/catalog.py, verification tiers, and the operator handoff for servers without Dynamic Client Registration.
npx skills add https://github.com/PostHog/posthog --skill adding-mcp-store-servers
The MCP store catalog is code: one entry in products/mcp_store/backend/catalog.py per server.
On deploy, sync_mcp_server_templates upserts entries into MCPServerTemplate rows in every environment — there are no data migrations, no icon assets, and no manual admin steps for most servers.
Adding a server is a small PR to that file.
products/mcp_store/README.md — the catalog pipeline, sync semantics, and operator runbookproducts/mcp_store/backend/catalog.py — existing entries; match their tone and shapeproducts/mcp_store/backend/probe.py — what the probe verifies and what passed_activation_gate meansCheck the vendor's docs (search "<vendor> MCP server"); most publish a hosted endpoint like https://mcp.<vendor>.com/mcp.
Only hosted (remote) MCP servers belong in the catalog — local/stdio servers do not.
Cross-check public MCP registries if the docs are unclear.
DEBUG=1 python manage.py probe_mcp_server https://mcp.example.com/mcp
The JSON verdict tells you everything the entry needs:
speaks_mcp: false → the probe could not verify MCP. Stop and re-research — with one exception: reachable: true plus the error "Initialize was rejected and no OAuth metadata was discovered" is the auth-walled API-key case (last bullet below).auth_flavor: "oauth_dcr" with passed_activation_gate: true → OAuth with Dynamic Client Registration. The entry is auth_type="oauth" and will activate automatically on merge.auth_flavor: "oauth_shared" → OAuth without DCR. The entry is auth_type="oauth" but ships inactive; an operator must register an OAuth app with the vendor and paste credentials in Django admin (see the operator checklist below — include it in your PR description).auth_flavor: "open" → the handshake completed without credentials. auth_type="api_key"; activates automatically on merge.auth_flavor: "api_key_or_unknown" with reachable: true → an auth-walled API-key server; a bare 401/403 gives the probe no MCP evidence. auth_type="api_key", but the entry ships inactive — verify it with a real install (Gate B) before adding it, and note in the PR that an operator flips it active in Django admin per environment (users bring their own key; nothing to provision).catalog.py, alphabetically by name:name — the vendor's own casing ("PagerDuty", not "Pagerduty").description — one sentence, sentence case, verb-first, matching the existing entries ("Manage Linear issues, projects, and team workflows."). No marketing copy.category — the closest of business / data / design / dev / infra / productivity.icon_domain — the vendor's primary brand domain (linear.app, not mcp.linear.app). Verify logo.dev has it: GET /api/projects/@current/hog_functions/icons/?query=<vendor> from a dev session, or check https://img.logo.dev/<domain> renders a real logo.docs_url — the vendor's MCP docs page when they have one.The probe covers everything up to the OAuth consent screen.
If you have an account with the vendor, complete one real install in local dev: run the stack, install the server from the store UI, finish the OAuth flow (or paste an API key), and confirm the tool list populates.
Record the verification tier in the PR description:
hogli test products/mcp_store/backend/test/test_catalog_sync.py
test_catalog_entries_are_valid catches malformed entries (bad category, duplicate URL, unnormalized icon_domain) before they hit production.
mcp-store/-prefixed branch (e.g. mcp-store/add-pagerduty), with a feat(mcp-store) title: feat(mcp-store): add <name> to the MCP server catalog.State the probe verdict and verification tier in the description.
For oauth_shared servers, include the operator checklist so activation isn't forgotten.
This server does not support Dynamic Client Registration, so it ships inactive. To activate (per environment, US and EU):
- [ ] Register an OAuth app in the vendor's developer console
- [ ] Redirect URI: `https://us.posthog.com/api/mcp_store/oauth_redirect/` (and the EU equivalent)
- [ ] Paste client ID + secret into Django admin → MCP server templates → <name>
- [ ] OAuth metadata was auto-discovered by the sync; run the "Discover metadata" admin action only if it's empty
- [ ] Tick "is active"
is_active, oauth_credentials, or oauth_metadata expectations into the catalog — those are operational state owned by the row, not by code.icon_key values — icons resolve from icon_domain via logo.dev at render time.> FHIR REST endpoints (Patient, Observation, Encounter, Condition, MedicationRequest), (2) Validating FHIR resources and returning proper HTTP status codes and error responses, (3) Implementing SMART on FHIR authorization and OAuth scopes, (4) Working with Bundles, transactions, batch operations, or search pagination. Covers FHIR R4 resource structures, required fields, value sets (status codes, gender, intent), coding systems (LOINC, SNOMED, RxNorm, ICD-10), and OperationOutcome error handling.
Interact with ClawDirect, a directory of social web experiences for AI agents. Use this skill to browse the directory, like entries, or add new sites. Requires ATXP authentication for MCP tool calls. Triggers: browsing agent-oriented websites, discovering social platforms for agents, liking/voting on directory entries, or submitting new agent-facing sites to ClawDirect.
Shared audit integrity framework for all AppSec agents — enforces output quality, intellectual honesty, and continuous improvement through anti-rationalization guards, self-critique loops, retry protocols, non-negotiable behaviors, self-reflection quality gates (1-10 scoring, ≥8 threshold), and a self-learning system with lesson/memory governance for security analysis agents.
Opt out of the OneCLI gateway and supply Anthropic credentials from .env instead. For users who want simple .env-based credential management without the OneCLI agent vault. Reads the API key or OAuth token from .env and injects it into the container's API requests.
Cross-product Zoom reference skill. Use after the workflow is clear when you need shared platform guidance, app-model comparisons, authentication context, scopes, marketplace considerations, or API-vs-MCP routing.
>- Static source-code vulnerability scan. Reads a target directory (and THREAT_MODEL.md if present), spawns parallel review subagents per focus area, and writes VULN-FINDINGS.json + .md for /triage to consume. Read-only — no building, running, or network. For execution-verified crashes, use vuln-pipeline instead. Use when asked to "scan for vulns", "review this code for security issues", "find bugs in <dir>", or as the step between /threat-model and /triage.
Hunt Session Management vulnerabilities — session fixation (no regeneration on login), insufficient invalidation on logout / password-change / email-change, predictable or low-entropy session IDs, JWT-as-session with no exp/revocation, refresh-token rotation/reuse-detection gaps, OAuth/SSO session linkage, device-bound-session (DBSC) downgrade, and cookie attribute issues (Secure/HttpOnly/SameSite/__Host-). Validate with TWO real sessions (attacker A + victim B), body-diff every 200, and OOB confirmation for theft chains. Medium to Critical (fixation→admin hijack, no-invalidation→persistent ATO).
> Use this skill when the user is doing hands-on DOCA AES-GCM work on a BlueField DPU or ConnectX NIC — configuring `doca_aes_gcm_task_encrypt` / `_task_decrypt`, querying `doca_aes_gcm_cap_*` for per-key-type (only `DOCA_AES_GCM_KEY_128` / `_256` — AES-192 not supported) and per-task support, sizing plaintext against the max-buf cap, setting source / destination mmap permissions, validating with a NIST GCMVS or RFC 5288 vector, or debugging DOCA_ERROR_* including the security-critical tag-verification-failed outcome on decrypt. Trigger even when the user does not explicitly mention "DOCA AES-GCM" or IO_FAILED", "auth tag isn't verifying", "NOT_PERMITTED on my encrypt buffer", "is AES-192-GCM on this BlueField" (no), or "encrypted record came back tampered". Refuse and route elsewhere for non-GCM AES modes (CBC / CTR / XTS — CPU OpenSSL), key management (KMS / HSM / rotation), SHA (doca-sha), or general AEAD background.
Take posthog/adding-mcp-store-servers from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.