Title: Add PulseTel - Project telemetry MCP server
Body:
PulseTel — Real-time project telemetry for AI agents
npm: pulsetel-cli
GitHub: siongyuen/pulsetel
Website: sycheah.com/pulsetel
Overview
PulseTel gives AI agents instant project health assessment — CI status, dependency vulnerabilities, endpoint health, issue triage, coverage, deploy status, and Sentry error tracking — in a single call with prioritised, actionable recommendations.
Unlike raw API wrappers (GitHub MCP) or paid platforms (Datadog), PulseTel returns agent-native responses: every tool includes actionable, severity, confidence, and context fields. The pulsetel_recommend tool returns a ranked action list so agents know exactly what to fix first.
MCP Tools (12)
| Tool |
Description |
| pulsetel_check |
Full health check with optional trends |
| pulsetel_quick |
Fast triage (~2s) — skips deps and coverage |
| pulsetel_ci |
CI flakiness score + trend |
| pulsetel_health |
Endpoint latency + baseline comparison |
| pulsetel_deps |
Vulnerability + outdated count |
| pulsetel_summary |
Summary + top anomalies + overall trend |
| pulsetel_trends |
Direction, delta, velocity per check |
| pulsetel_anomalies |
2σ anomaly detection |
| pulsetel_metrics |
History + trends + current values |
| pulsetel_recommend |
Prioritised action items by severity |
| pulsetel_status |
Lightweight health ping (sub-10ms, no API calls) |
| pulsetel_sentry |
Sentry error tracking — unresolved issues, error rates, affected users |
Transport
- stdio (Claude Desktop, Cursor, Smithery)
- HTTP (REST API on port 3000)
Configuration
{
"mcpServers": {
"pulsetel": {
"command": "npx",
"args": ["-y", "pulsetel-cli", "mcp-stdio"]
}
}
}
Sentry Integration
Configure Sentry in .pulsetel.yml:
sentry:
organization: my-org
project: my-project
# token: prefer SENTRY_TOKEN env var instead
The pulsetel_sentry tool returns unresolved issue counts grouped by severity, top 5 issues by frequency, affected user counts, and release attribution — all with actionable recommendations.
OpenTelemetry Integration
PulseTel exports project health as OTel-compliant telemetry:
• Traces — Root span per check with child spans per module
• Metrics — Health scores, anomaly counters, dependency counts
• Logs — Anomaly events, degrading trends, flaky CI detection
Want me to use the GitHub CLI to update it directly?
Title: Add PulseTel - Project telemetry MCP server
Body:
PulseTel — Real-time project telemetry for AI agents
npm: pulsetel-cli
GitHub: siongyuen/pulsetel
Website: sycheah.com/pulsetel
Overview
PulseTel gives AI agents instant project health assessment — CI status, dependency vulnerabilities, endpoint health, issue triage, coverage, deploy status, and Sentry error tracking — in a single call with prioritised, actionable recommendations.
Unlike raw API wrappers (GitHub MCP) or paid platforms (Datadog), PulseTel returns agent-native responses: every tool includes actionable, severity, confidence, and context fields. The pulsetel_recommend tool returns a ranked action list so agents know exactly what to fix first.
MCP Tools (12)
Transport
Configuration
{ "mcpServers": { "pulsetel": { "command": "npx", "args": ["-y", "pulsetel-cli", "mcp-stdio"] } } } Sentry Integration Configure Sentry in .pulsetel.yml: sentry: organization: my-org project: my-project # token: prefer SENTRY_TOKEN env var instead The pulsetel_sentry tool returns unresolved issue counts grouped by severity, top 5 issues by frequency, affected user counts, and release attribution — all with actionable recommendations. OpenTelemetry Integration PulseTel exports project health as OTel-compliant telemetry: • Traces — Root span per check with child spans per module • Metrics — Health scores, anomaly counters, dependency counts • Logs — Anomaly events, degrading trends, flaky CI detection Want me to use the GitHub CLI to update it directly?