Regression Watch
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
mcp-regression-watch
MCP server for the Agent Update Regression Watch - deploy-triggered regression testing: post-update behavior-diffing against baselines, drift alerts, retained regression-comparison reports. Scheduled benchmarks catch planned drift; this catches the unplanned kind.
Pricing: $29-99/mo (rides the proofspan tier).
Tools
run_regression_check
Run a deploy-triggered regression check against the persisted baseline.
Parameter | Type | Default | Description |
| string | required | Service being checked (e.g. mcp-agent-sla) |
| string | required | Version under test (e.g. v2.1.0) |
| integer | - | Optional seed for reproducible simulation |
| boolean |
| Persist the run to state |
Returns per-case results with baseline value, current value, delta, pct drift, threshold, and status. Overall pass/fail plus breach list. Persists the raw run log (customer-owned exit material per two-column IP rule).
get_drift_report
Aggregate persisted run history into a drift report.
Parameter | Type | Default | Description |
| string | required | Service to report on |
| integer | 30 | Look-back window in days (1-365) |
Returns drift trends per metric over time, breach counts, worst-offending version, drift heat summary, and alert list. All values computed from actual stored runs via Python statistics.
get_comparison_report
Retained comparison report between two versions.
Parameter | Type | Default | Description |
| string | required | Service to compare |
| string | required | Baseline version |
| string | required | Candidate version |
Returns per-case side-by-side table, diff summary (improved/regressed/unchanged/added/removed), severity-ranked regression list, and content hash.
set_drift_threshold
Persist a per-metric drift threshold.
Parameter | Type | Default | Description |
| string enum | required | Metric (latency_p95_ms, tool_call_accuracy, hallucination_rate, json_schema_validity, token_cost, refusal_precision, retry_rate, throughput_rps, error_rate) |
| number | required | Drift threshold as percentage of baseline (0.1-100.0) |
| boolean |
| Persist to state |
Related MCP server: deploy-check
Installation
pip install -r requirements.txtUsage
python main.pyState persisted in ./state/regression_state.json. Override with MCP_REGRESSION_WATCH_STATE_DIR.
Connecting to an MCP client
{
"mcpServers": {
"regression-watch": {
"command": "python3",
"args": ["/tmp/v3-services/mcp-regression-watch/main.py"]
}
}
}License
MIT
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
Synthetic checks, nightly regression replay and model-drift alerts for AI agents
Deterministic agent checks: manifests, schemas, receipts, OpenAPI drift, DNS, TLS, headers.
121QA platform for agents: coverage signals, in-repo test plans, verified tests and release governance.
Test analytics for AI agents: test history, flaky tests, failure clusters, coverage.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to prove that previously working behavior still works after code changes by running the product through the same steps twice and surfacing only unintended differences as actionable findings.19 npmMIT
- AlicenseAqualityBmaintenanceEnables agents to verify that a deploy actually worked by checking pages, links, and comparing environments, catching blank 200 responses, broken links, stray noindex, and staging that never updated.3MIT
- AlicenseNot gradedqualityCmaintenanceEnables ingesting caller-supplied agent execution traces and comparing them against signed, versioned baselines of normal behaviour to surface silent drift. Reports severity with per-dimension attribution, a replayable exemplar episode, and visible suppression of declared maintenance windows or known change markers.1MIT
- AlicenseNot gradedqualityBmaintenanceEnables MCP clients to check whether third-party API responses have drifted, inspect expected schemas, review change history, list monitored endpoints, and trigger checks before agents trust an API.278 PyPI1MIT