Pulse ServersBLOCK
MCP (Model Context Protocol) Servers authored and maintained by the PulseMCP team. We build reliable servers thoughtfully designed specifically for MCP Client-powered workflows.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
We build reliable servers thoughtfully designed for MCP Client-powered workflows.
Design principles
You can have confidence that any Pulse-branded MCP server was built with these north stars in mind:
- Purpose-built. LLM-powered MCP clients offer uniquely new user interaction patterns that necessitate a different layer of abstraction than the REST API's from a pre-AI era. We thoughtfully scope drawing lines like building a new server, versus incorporating a new feature in another server. Or deciding where one slew of REST API endpoints might be better packaged as a single Tool call. And more.
- Easy set up. Many MCP servers die before ever getting a chance to be used. We offer guides and a frustration-free experience to get going with our MCP servers inside your favorite MCP clients.
- Time savings. By minimizing the number of tool chain steps or conversational turns you need to accomplish a task, our MCP servers will save you (and your agents) time waiting for a task to be completed.
- Inference cost savings. By minimizing the number of tokens you need to consume to a accomplish a task, our MCP servers will save you on your LLM inference bills.
- Reliability. You should have confidence that you can deploy our servers in a production application serving mass market consumers or business clients.
- Future-proof. We sit on the bleeding edge of the MCP specification, working to push the ecosystem forward. As such, you can be sure that if you commit to baking our server into your workflow, it will self-improve over time to take advantage of the latest and greatest MCP features.
Servers Available
Productionized Servers
These are PulseMCP-branded servers that we intend to maintain indefinitely as our own offerings.
The MCP Registry column controls whether the server is published to the official MCP Registry (com.pulsemcp/) by the `p
de59a2a5e039OBSERVED · 2026-10-07Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add agent-orchestrator-mcp-server --env AGENT_ORCHESTRATOR_API_KEY=${AGENT_ORCHESTRATOR_API_KEY} -- npx -y [email protected]Exposed tools (193)
127 read · 56 write · 10 destructive. Blast radius: 10 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
AGENT_ORCHESTRATOR_API_KEY | read | API key for authentication with the Agent Orchestrator |
AGENT_ORCHESTRATOR_BASE_URL | read | Base URL for the Agent Orchestrator API |
ALLOWED_AGENT_ROOTS | write | Comma-separated list of allowed agent root names. When set, only these agent roots are shown in get_configs and allowed in start_session (with their exact default MCP servers only) |
APPSIGNAL_API_KEY | read | AppSignal API key for authentication |
APPSIGNAL_APP_ID | read | AppSignal application ID (optional - can be selected dynamically) |
AWS_ACCESS_KEY_ID | write | AWS access key ID (uses default credential chain if not set) |
AWS_DEFAULT_REGION | read | Alternative region variable (fallback) |
AWS_ENDPOINT_URL | read | Custom S3 endpoint URL (for S3-compatible services like MinIO) |
AWS_REGION | read | AWS region for S3 operations |
AWS_SECRET_ACCESS_KEY | write | AWS secret access key (uses default credential chain if not set) |
BROWSER_PERMISSIONS | write | Comma-separated list of browser permissions to grant. If not set, ALL permissions are granted. |
Cleanliness | read | Reviews mentioning cleanliness |
DANGEROUSLY_SKIP_ELICITATIONS | write | Set to |
DISABLE_DOCKER_CLI_TOOLS | write | Disable Docker CLI-based tools (registry tools) |
DYNAMODB_DISABLED_TOOLS | write | Comma-separated blacklist of specific tools to disable |
DYNAMODB_ENABLED_TOOLS | write | Comma-separated whitelist of specific tools to enable |
DYNAMODB_ENABLED_TOOL_GROUPS | write | Comma-separated list of tool groups to enable (readonly, readwrite, admin) |
DYNAMODB_ENDPOINT | read | Custom DynamoDB endpoint (for local DynamoDB or LocalStack) |
ENABLED_TOOLGROUPS | write | Comma-separated list of tool groups to enable (readonly,write,admin,apps,machines,logs,ssh,images,registry) |
FETCHPET_DOWNLOAD_DIR | write | Directory to save downloaded documents (EOB, invoices) |
FETCHPET_PASSWORD | read | Your Fetch Pet account password |
FETCHPET_USERNAME | read | Your Fetch Pet account email address |
FLY_IO_API_TOKEN | read | API token for Fly.io authentication |
GCS_BUCKET | read | Constrain all operations to a single bucket (hides bucket-level tools) |
GCS_DISABLED_TOOLS | write | Comma-separated list of specific tools to disable |
GCS_ENABLED_TOOLGROUPS | destructive | Comma-separated list of tool groups to enable (readonly, readwrite, delete) |
GCS_ENABLED_TOOLS | write | Comma-separated list of specific tools to enable (overrides groups) |
GCS_MAKE_PUBLIC | read | Whether to make uploaded files publicly accessible by default |
GCS_PROJECT_ID | read | Google Cloud project ID |
GCS_ROOT_PATH | read | Root path prefix - server cannot access above this path |
GCS_SERVICE_ACCOUNT_KEY_FILE | read | Path to service account key JSON file |
GCS_SERVICE_ACCOUNT_KEY_JSON | read | Service account key JSON contents (inline) |
GOOD_EGGS_PASSWORD | read | Your Good Eggs account password |
GOOD_EGGS_USERNAME | read | Your Good Eggs account email address |
Goals | read | Available session completion criteria for use with start_session. |
HATCHBOX_ACCOUNT_ID | read | Your Hatchbox account ID |
HATCHBOX_API_KEY | read | Hatchbox API key for authentication |
HATCHBOX_APP_ID | read | Your Hatchbox application ID |
HATCHBOX_DEPLOY_KEY | write | Deployment webhook key |
HEADLESS | write | Run browser in headless mode (true/false) |
HEALTH_CHECK_TIMEOUT | read | Health check timeout in milliseconds |
IGNORE_HTTPS_ERRORS | write | Ignore HTTPS certificate errors (true/false). Set to false for stricter security. |
LANGFUSE_BASE_URL | read | Langfuse API base URL |
LANGFUSE_PUBLIC_KEY | read | Langfuse public key for API authentication |
LANGFUSE_SECRET_KEY | read | Langfuse secret key for API authentication |
Location | read | Reviews mentioning location |
OP_ELICITATION_READ | read | Prompt before revealing credentials (true/false) |
OP_ELICITATION_WRITE | write | Prompt before creating/sharing items (true/false) |
OP_SERVICE_ACCOUNT_TOKEN | read | 1Password service account token for authentication |
OP_WHITELISTED_ITEMS | read | Comma-separated item titles or IDs that bypass read elicitation |
PROCTOR_API_KEY | read | API key for PulseMCP Proctor API authentication |
PROCTOR_API_URL | read | Base URL for Proctor API (default: https://admin.pulsemcp.com) |
PROXY_BYPASS | read | Comma-separated list of hosts to bypass proxy |
PROXY_PASSWORD | read | Proxy authentication password |
PROXY_URL | read | Proxy server URL (e.g., http://proxy.example.com:8080) |
PROXY_USERNAME | read | Proxy authentication username |
PULSEMCP_ADMIN_API_KEY | read | API key for PulseMCP admin API authentication |
S3_BUCKET | read | Constrain all operations to a single bucket (hides bucket-level tools) |
S3_DISABLED_TOOLS | write | Comma-separated list of specific tools to disable |
S3_ENABLED_TOOLGROUPS | write | Comma-separated list of tool groups to enable (readonly, readwrite) |
S3_ENABLED_TOOLS | write | Comma-separated list of specific tools to enable (overrides groups) |
S3_FORCE_PATH_STYLE | destructive | Use path-style addressing instead of virtual-hosted-style (required for MinIO and some S3-compatible services) |
SKIP_HEALTH_CHECKS | write | Skip API connectivity check at startup (set to |
SLACK_BOT_TOKEN | read | Slack Bot User OAuth Token (starts with xoxb-) |
SSH_AUTH_SOCK | read | Path to SSH agent socket for agent-based authentication (recommended for passphrase-protected keys) |
SSH_HOST | read | Hostname or IP address of the remote SSH server |
SSH_PASSPHRASE | read | Passphrase for encrypted private key (only if not using SSH agent) |
SSH_PORT | read | SSH port number |
SSH_PRIVATE_KEY_PATH | read | Path to SSH private key file (alternative to SSH agent) |
SSH_TIMEOUT | read | Connection timeout in milliseconds |
SSH_USERNAME | read | Username for SSH authentication |
STEALTH_LOCALE | read | Custom locale for Accept-Language header (e.g., en-US,en) |
STEALTH_MASK_LINUX | read | Mask Linux platform as Windows in stealth mode (true/false) |
STEALTH_MODE | write | Enable stealth mode to bypass anti-bot protection (true/false) |
STEALTH_USER_AGENT | read | Custom User-Agent string for stealth mode (overrides default) |
Service | read | Reviews mentioning service |
TAILSCALE_API_URL | read | Base URL for the Tailscale API (default: https://api.tailscale.com/api/v2) |
TAILSCALE_TAILNET | read | Tailnet to operate on (default: |
TIMEOUT | read | Default timeout for browser operations in milliseconds |
TOOL_GROUPS | write | Comma-separated list of tool groups to enable (sessions,sessions_readonly,notifications,notifications_readonly,triggers,triggers_readonly,health,health_readonly,self_session) |
TWIST_BEARER_TOKEN | read | Twist API bearer token for authentication |
TWIST_WORKSPACE_ID | read | Twist workspace ID to operate within |
VERCEL_ENABLED_TOOLGROUPS | write | Comma-separated list of tool groups to enable (readonly,readwrite) |
VERCEL_TEAM_ID | read | Team ID for team-scoped operations |
VERCEL_TEAM_SLUG | read | Team URL slug for team-scoped operations |
VERCEL_TOKEN | read | Vercel API token for authentication |
add_official_mirror_to_regular_queue | write | Add an official mirror queue entry to the regular MCP implementation queue. This creates a draft MCP implementation from the mirror data. This is an async operation. This action: 1. Creates a new draft MCP implementation using the mirror |
add_servers_to_tenant | write | Add MCP servers to a tenant |
authorize_device | read | Authorize or deauthorize a device. Sets the authorization state of a device. Relevant for tailnets that require device authorization — an unauthorized device cannot connect. **Returns:** - Confirmation of the new authorization state **Use cases:** - Approve a newly-joined device - Revoke a device |
chat | read | Sends a message to the subagent and waits for the response. The subagent will process the message using its installed MCP servers and return the result. This is the primary way to interact with the configured subagent. Example response: { |
cleanup_good_jobs | read | Clean up old GoodJob background jobs by deleting them. Removes jobs matching the specified criteria to free up database space. Example request: { |
copy_object | read | Copy an object within GCS (same bucket or across buckets). Copies an object from one location to another. Can be used to rename objects or move them between buckets. Example response: { |
create_api_key | write | Create an API key for a specified tenant. Returns the raw key value, which is only available at creation time. Example request: { |
create_bucket | write | Create a new GCS bucket. Creates an empty bucket with the specified name in the given location. Example response: { |
create_calendar_event | write | Creates a new event in Google Calendar. |
create_mcp_json | write | Create a new MCP JSON configuration. MCP JSONs define how MCP servers should be launched. Example request: { |
create_oauth_client | write | Create a Tailscale OAuth client (client_id + client_secret). Mints a non-interactive credential for API automation. Unlike a device auth key (create_auth_key) — which only enrolls devices — an OAuth client can be exchanged for short-lived access tokens that call \ |
create_redirect | write | Create a new URL redirect entry. Redirects manage URL routing on the PulseMCP website. Example request: { |
create_tenant | write | Create a new tenant for sub-registry provisioning. Example request: { |
create_unofficial_mirror | write | Create a new unofficial mirror entry. Unofficial mirrors represent community-submitted MCP server definitions. The server_json parameter accepts server.json content directly and automatically wraps it in a { |
delete_bucket | destructive | Delete a GCS bucket. Permanently removes the specified bucket. The bucket must be empty before deletion. Example response: { |
delete_calendar_event | destructive | Deletes an event from Google Calendar. |
delete_mcp_json | destructive | Delete an MCP JSON configuration by its ID. This action is irreversible. Use cases: - Remove obsolete configurations - Clean up test data - Remove incorrectly created configurations |
delete_object | destructive | Delete an object from GCS. Permanently removes the specified object from the bucket. Example response: { |
delete_redirect | destructive | Delete a URL redirect by its ID. This action is irreversible. Use cases: - Remove duplicate or outdated redirects - Clean up test data - Remove redirects that were incorrectly created |
delete_unofficial_mirror | destructive | Delete an unofficial mirror by its ID. This action is irreversible. Use cases: - Remove duplicate or outdated unofficial mirrors - Clean up test data - Remove mirrors that were incorrectly created |
discard_good_job | read | Discard a GoodJob background job. Marks the job as discarded so it will not be retried. Use cases: - Discard a job that is no longer needed - Stop a failing job from being retried - Clean up jobs that are stuck or obsolete |
download_object | read | Download a single object to a local file path, writing raw bytes (binary-safe). Unlike get_object (which returns content inline as UTF-8 text and is lossy for binary or large files), this tool streams the object to disk and returns the local path. Example response: { |
draft_newsletter_post | write | Create a new draft newsletter post in the PulseMCP CMS. All posts are created as drafts initially and can be published later through the CMS interface. This tool handles all content fields, SEO metadata, and associations with MCP servers/clients. Example successful response: { |
engineering | read | Engineering team channel |
find_airport_code | read | Look up airport IATA codes by city name, airport name, or partial code. Use this to find the correct airport code before calling search_flights or get_date_grid. Returns matching airports with their IATA code, full name, city, and country. |
find_servers | read | Analyzes your trusted servers list to determine which MCP servers are relevant for the given task. This tool uses intelligent analysis to match task requirements with server capabilities, ensuring the subagent only gets the tools it needs. Example response: { |
force_trigger_good_job_cron | destructive | Force trigger a GoodJob cron schedule immediately. Enqueues a new job for the specified cron schedule regardless of its normal schedule. Use cases: - Manually trigger a cron job for testing - Force an immediate run of a periodic job - Re-run a cron job that was missed |
general | read | General discussion |
getEnvVar | read | Retrieve a specific environment variable from your Rails application running on Hatchbox via SSH. This tool connects to your server and reads the requested environment variable directly from the running puma process. Example response: RAILS_ENV=production Or if not found: Environment variable |
getEnvVars | read | Retrieve all environment variables from your Rails application running on Hatchbox via SSH. This tool connects to your server and reads environment variables directly from the running puma process, providing a complete view of your application |
get_anomaly_incident | read | Retrieve detailed information about a specific anomaly incident in your AppSignal application. Anomaly incidents are automatically detected unusual patterns in your application |
get_anomaly_incidents | read | Retrieve a list of anomaly incidents detected by AppSignal |
get_calendar_event | read | Retrieves detailed information about a specific calendar event by ID. |
get_discovered_url_stats | read | Get summary statistics for discovered URLs. Returns counts of pending URLs and today |
get_draft_mcp_implementations | read | Retrieve a paginated list of draft MCP implementations from the PulseMCP Admin panel with all associated objects. Returns formatted markdown with implementation details, metadata, and linked resources. This tool returns the same data shown in the Admin panel sidebar under |
get_good_job | read | Retrieve a single GoodJob background job by its ID. Returns detailed information including job class, queue, status, and error details. Example response: { |
get_good_job_queue_statistics | read | Retrieve aggregate statistics for GoodJob background jobs. Returns counts of jobs grouped by status. Example response: { |
get_key | read | Get metadata about a single auth key or API access token. Returns the metadata for one key. Secret key material is NOT returned. **Returns:** - The key object: id, creation/expiry times, capabilities, and description **Use cases:** - Inspect a key |
get_mcp_json | read | Retrieve a single MCP JSON configuration by its ID. Returns the full configuration value. Example response: { |
get_mcp_jsons | read | Retrieve a paginated list of MCP JSON configurations. MCP JSONs contain server configuration templates that can be used to set up MCP servers. Example response: { |
get_moz_backlinks | read | Fetch live backlink data from the MOZ API. Returns source pages, anchor text, and domain authority for backlinks pointing to a given URL. Example response: { |
get_moz_metrics | read | Fetch live URL metrics from the MOZ API. Returns page authority, domain authority, spam score, and link counts for a given URL. Example response: { |
get_moz_stored_metrics | read | List stored/historical MOZ data for a server |
get_object | read | Get the contents of an object from GCS. Returns the object content as text along with metadata. Example response: { |
get_official_mirror | read | Retrieve a single official mirror by its ID or name. Returns detailed information including the full JSON data. Example response: { |
get_official_mirror_queue_items | read | Retrieve a paginated list of official mirror queue entries from the PulseMCP Admin panel. Returns formatted markdown with queue item details, status, and linked server information. The official mirror queue contains entries from the MCP Registry |
get_official_mirrors | read | Retrieve a paginated list of official mirrors from the MCP Registry. These are mirrors that come from the official MCP Registry |
get_perf_incident | read | Retrieve details about a specific performance incident from AppSignal. Performance incidents represent specific performance bottlenecks like slow endpoints, database queries, or external API calls. This tool provides detailed information about a single performance issue. Example response: { |
get_redirect | read | Retrieve a single URL redirect by its ID. Returns detailed information including the from/to paths and status. Example response: { |
get_redirects | read | Retrieve a paginated list of URL redirects from the PulseMCP Admin API. Redirects manage URL routing on the PulseMCP website. Example response: { |
get_server | read | Get detailed information about a specific MCP server from the PulseMCP Sub-Registry. Returns the server\ |
get_server_capabilities | read | Retrieves capabilities and descriptions for specified MCP servers. This tool helps understand what each server can do without installing them, useful for making informed decisions about server selection. Example response: { |
get_tenant | read | Retrieve a single tenant by its ID or slug. Returns detailed information including enrichments. Example response: { |
get_tenants | read | Retrieve a paginated list of tenants. Tenants represent organizations or users that have access to the PulseMCP platform. Example response: { |
get_thread | read | Retrieve a complete thread conversation including all messages. This tool fetches the full context of a discussion, showing the thread metadata and every message posted in chronological order. Essential for understanding the complete conversation history. Example response: Thread: |
get_unofficial_mirror | read | Retrieve a single unofficial mirror by its ID or name. Returns detailed information including the full JSON data. Example response: { |
get_unofficial_mirrors | read | Retrieve a paginated list of unofficial mirrors from the PulseMCP Admin API. Unofficial mirrors are community-submitted MCP server definitions that contain metadata like name, version, and configuration data. Example response: { |
github-development | read | Interact with GitHub repositories, issues, and pull requests |
head_bucket | read | Check if a GCS bucket exists and is accessible. Returns whether the bucket exists and the caller has permission to access it. Example response: { |
init_agent | read | Initializes a Claude Code subagent with a custom system prompt and working directory. This tool creates a new Claude Code instance in an isolated environment. The agent |
inspect_transcript | read | Retrieves the subagent |
install_servers | write | Installs and configures MCP servers on the subagent using your servers.json configurations. This tool updates the subagent |
list_buckets | read | List all GCS buckets in the Google Cloud project. Returns a list of bucket names and their creation dates. Example response: { |
list_calendar_events | read | Lists events from a Google Calendar within an optional time range. |
list_calendars | read | Lists all calendars available to the authenticated user. |
list_discovered_urls | read | List discovered URLs filtered by processing status, with pagination. Returns URLs ordered by created_at ASC (oldest first). Example response: { |
list_good_job_cron_schedules | read | Retrieve a list of all GoodJob cron schedules. Returns configured recurring job schedules with their cron expressions and next run times. Example response: [ { |
list_good_job_processes | read | Retrieve a list of all active GoodJob worker processes. Returns information about running worker processes including hostname, PID, and queue assignments. Example response: [ { |
list_good_jobs | read | Retrieve a paginated list of GoodJob background jobs from the PulseMCP Admin API. Example response: { |
list_objects | read | List objects in a GCS bucket with optional prefix filtering and pagination. Example response: { |
list_servers | read | Browse MCP servers from the PulseMCP Sub-Registry. Returns a paginated list of active and deprecated servers with their names, descriptions, and metadata. Deleted servers are excluded by default. Use search to filter by name or description. Use cursor for pagination through large result sets. |
list_tenant_servers | read | List a tenant |
mark_discovered_url_processed | read | Mark a single discovered URL as processed after the agent has handled it. Sets processed_at to the current timestamp and stores the result for audit trail. Idempotent — calling on an already-processed URL updates the fields. Use cases: - Mark a URL as |
mcp-servers | read | PulseMCP MCP servers monorepo |
mock-channel | read | Mock channel |
put_object | write | Upload or update an object in GCS. Creates a new object or overwrites an existing one with the provided content. Example response: { |
query_calendar_freebusy | read | Queries free/busy information for one or more calendars within a specified time range. |
recache_mcp_server | read | Refresh the cache for a specific MCP server. Clears and warms the show page, card fragments across all sort variations, canonical URLs, and parent pages. Example request: { |
remove_servers_from_tenant | destructive | Remove MCP servers from a tenant |
reschedule_good_job | read | Reschedule a GoodJob background job to a new time. Changes the scheduled_at time for the job. Example request: { |
retry_good_job | read | Retry a failed or discarded GoodJob background job. Re-enqueues the job for processing. Use cases: - Retry a job that failed due to a transient error - Re-process a discarded job after fixing the underlying issue |
run_exam | write | Execute a Proctor exam against an MCP server. Runs the specified exam using the provided runtime and MCP configuration. The exam tests the MCP server |
save_mcp_implementation | write | Create or update an MCP implementation. This tool replicates the |
save_results_for_mirror | write | Save proctor exam results for an unofficial mirror. Pass the \ |
search_flights | read | Search for flights on Google Flights with full filtering and pagination. Returns structured flight data including prices, airlines, times, durations, stops, individual flight segments with aircraft type and legroom, fare brand (e.g. |
search_logs | read | Search through application logs in AppSignal with powerful filtering capabilities. This tool allows you to query logs by content, filter by severity levels, and retrieve recent log entries matching your criteria. It |
send_impl_posted_notif | write | Send an email notification to the submitter of an MCP implementation that has been published to PulseMCP. This tool sends the actual notification email (not just a draft) to inform submitters that their MCP implementation is now live on PulseMCP. It replicates and extends the |
set_device_tags | write | Set the tags on a device. Replaces the device |
set_github_repository_classification | write | Set the \ |
set_known_missing_init_tools_list | write | Update the \ |
slack | write | Send and receive messages in Slack workspaces |
ssh_connection_info | read | Get information about the configured SSH connection. **Returns:** JSON object with host, port, username, and authentication method details **Use cases:** - Verify SSH connection configuration - Debug connection issues - Confirm which server is being targeted |
ssh_download | read | Download a file from the remote SSH server to the local machine via SFTP. **Returns:** Success message with transfer details **Use cases:** - Retrieve log files for analysis - Download backups or exports - Fetch configuration files for review |
ssh_execute | write | Execute a command on the remote SSH server. **Returns:** JSON object with stdout, stderr, and exit code **Use cases:** - Run shell commands on remote servers - Check system status (uptime, disk space, processes) - Execute scripts or programs remotely - Manage services and applications |
ssh_upload | write | Upload a file from the local machine to the remote SSH server via SFTP. **Returns:** Success message with transfer details **Use cases:** - Deploy configuration files to servers - Upload scripts for remote execution - Transfer data files for processing |
stop_agent | write | Gracefully stops the running subagent. This tool ensures proper cleanup of resources and saves the final agent state before termination. Example response: { |
switch_tenant_id | read | Switch the active tenant ID used for all subsequent API requests. This is an admin tool for multi-tenant scenarios. Pass an empty string to clear the tenant ID. |
test-server-1 | read | A test server for integration testing |
test-server-2 | read | Another test server |
triggerDeploy | write | Trigger a new deployment for your Rails application on Hatchbox. This tool initiates a deployment using Hatchbox |
update_calendar_event | write | Updates an existing event in Google Calendar. |
update_mcp_json | write | Update an existing MCP JSON configuration by its ID. Only provided fields will be updated. Example request: { |
update_mcp_server | write | Update an MCP server |
update_newsletter_post | write | Update an existing newsletter post |
update_redirect | write | Update an existing URL redirect by its ID. Only provided fields will be updated. Example request: { |
update_unofficial_mirror | write | Update an existing unofficial mirror by its ID. Only provided fields will be updated. The server_json parameter accepts server.json content directly and automatically wraps it in a { |
upload_image | write | Upload an image file and attach it to a specific newsletter post. Images are stored in cloud storage organized by post, and the returned URL can be used in any of the post |
Trust audit
BLOCKgrade F · trust 42/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (7 observation(s))
- Network
- declared (6 observation(s))
- Shell
- declared (6 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- found
Findings (25)
await exec(tool.command);
token: 'integration-mock-token',
"GMAIL_SERVICE_ACCOUNT_PRIVATE_KEY": "-----BEGIN PRIVATE KEY-----\nMIIE...\n-----END PRIVATE KEY-----",
- `private_key`: The private key (including `-----BEGIN PRIVATE KEY-----` and `-----END PRIVATE KEY-----`)
export GCAL_SERVICE_ACCOUNT_PRIVATE_KEY="-----BEGIN PRIVATE KEY-----\nYour private key content\n-----END PRIVATE KEY-----"
"GCAL_SERVICE_ACCOUNT_PRIVATE_KEY": "-----BEGIN PRIVATE KEY-----\nYour private key\n-----END PRIVATE KEY-----",
"GOOGLE_DOCS_SERVICE_ACCOUNT_PRIVATE_KEY": "-----BEGIN PRIVATE KEY-----\nMIIE...\n-----END PRIVATE KEY-----",
SLACK_BOT_TOKEN: xoxb-test-token
SLACK_BOT_TOKEN: xoxb-test-token
"SLACK_BOT_TOKEN": "xoxb-your-token-here"
GCS_ENABLED_TOOLGROUPS, S3_FORCE_PATH_STYLE, delete_bucket, delete_calendar_event, delete_mcp_json, delete_object, delete_redirect, delete_unofficial_mirror, force_trigger_good_job_cron, remove_server
.prettierignore
.prettierrc.json
.eslintignore
.eslintignore
.eslintignore
CLAUDE.md
experimental/CLAUDE.md
libs/mcp-server-template/CLAUDE.md
libs/mcp-server-template/shared/src/example-client/CLAUDE.md
import type { AnomalyIncidentData } from '../../types.js';import type { App } from '../../graphql/types.js';import type { AnomalyIncidentData, IncidentListResult } from '../../types.js';import type { IncidentListResult } from '../../types.js';import type { IncidentListResult } from '../../types.js';Gates applied: no_behavioural_pass.
de59a2a5e039full audit observations/trust-audit/mcp-server/pulsemcp__pulse-servers.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-07 | de59a2a5e039 | BLOCK | F | 42 | first audit |
Questions
What is the Pulse Servers MCP server?
MCP (Model Context Protocol) Servers authored and maintained by the PulseMCP team. We build reliable servers thoughtfully designed specifically for MCP Client-powered workflows.
What tools does Pulse Servers expose?
193 in total: 127 read-only, 56 that write, and 10 that can delete or overwrite (GCS_ENABLED_TOOLGROUPS, S3_FORCE_PATH_STYLE, delete_bucket, delete_calendar_event, delete_mcp_json). Every one is listed on this page with its risk.
Is Pulse Servers safe to connect to an agent?
No — not without reading the findings first. The audit graded it F (42/100) and found 1 critical or high issue in the source. Each one is listed on this page with the file and line it is on. Separately from the audit: 10 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does Pulse Servers need?
It reads AGENT_ORCHESTRATOR_API_KEY, APPSIGNAL_API_KEY, AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, BRIGHTDATA_API_KEY, FETCHPET_PASSWORD, FIRECRAWL_API_KEY, FLY_IO_API_TOKEN, GCAL_SERVICE_ACCOUNT_PRIVATE_KEY, GCS_PRIVATE_KEY, GCS_SERVICE_ACCOUNT_KEY_FILE and GCS_SERVICE_ACCOUNT_KEY_JSON from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does Pulse Servers run?
It speaks stdio and streamable-http, so it runs as a local process your client starts. It is published on npm as @pulsemcp/svg-tracer-mcp-server-shared at 0.1.0.
How current is this page?
The grade is for one exact copy of the source (de59a2a5e039), read on 2026-10-07. The repository is watched and re-audited when it changes.