Estimate software reliability growth
Estimate long-run software reliability growth with a power-law nonhomogeneous Poisson process, bootstrap trend evidence, and future incident exposure.
What it's for
Distinguishes durable reliability improvement from a constant or worsening incident baseline and turns the trend into forward incident risk.
What you give it
Every tool is a pure function: it computes from what you pass it and reads nothing on its own. Inputs split into evidence you supply from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.
| Field | Type | Role | Required |
|---|---|---|---|
| bootstrap_draws | integer ≥ 200, ≤ 20000 | Numerical control | Optional |
| confidence_level | number ≥ 0.8, ≤ 0.999 | Your calibration | Optional |
| horizons_after_observation_days | array of number ≥ 1 item | Evidence | Yes |
| max_detail_rows | integer ≥ 1, ≤ 5000 | Numerical control | Optional |
| seed | integer | Numerical control | Optional |
| systems | array of objects (3 fields) | Evidence | Yes |
Each systems
record
| Field | Type | Required |
|---|---|---|
| id | string (non-empty) | Yes |
| incident_times_days | array of number (≥ 8 items) | Yes |
| observation_end_days | number (> 0) | Yes |
{
"bootstrap_draws": 200,
"horizons_after_observation_days": [
30,
90
],
"seed": 3,
"systems": [
{
"id": "checkout-service",
"incident_times_days": [
0.05161955923550008,
0.2479796737906886,
0.5144496422295098,
0.8319499619482423,
1.1912910440166142,
1.5867851514329019,
2.0144752596476954,
2.471409216800811,
2.9552788473745992,
3.4642174931764935,
3.9966765162190416,
4.551345156042946,
5.1270959750477365,
5.722946279669972,
6.3380299764748,
6.971576501395341,
7.622894694490652,
8.291360224997495,
8.976405623597778,
9.677512267545591,
10.394203854139318,
11.126041026088336,
11.872616900689994,
12.633553316946166,
13.408497659346127,
14.197120149535808,
14.999111521118527,
15.814181010835721,
16.64205461303338,
17.48247355479738,
18.335192957262347,
19.199980654954977,
20.076616150051013,Truncated for display — the full payload is 116 lines.
What you get back
This is the actual output of running the example above — computed by the same function the platform calls, not an illustration.
{
"assumptions": [
"Incident arrivals follow a power-law nonhomogeneous Poisson process over a stable observation boundary with complete event capture.",
"The growth shape is constant over the window; major architecture, traffic, reporting, or severity-definition changes require stratification or a new epoch.",
"A shape below one means declining incident intensity and reliability growth; it does not prove that any particular engineering intervention caused the improvement.",
"This long-run baseline-trend model complements rather than replaces short-run self-exciting incident-cluster and dependency-cascade models."
],
"decision": "reliability_growth_evidence_available",
"method": "power_law_process_software_reliability_growth_v1",
"portfolio": {
"detail_rows_returned": 1,
"detail_rows_truncated": 0,
"deteriorating_systems": 0,
"improving_systems": 1,
"inconclusive_systems": 0,
"systems": 1
},
"systems": [
{
"current_incident_intensity_per_day": 0.702432,
"current_mean_time_between_incidents_days": 1.423626,
"fit": {
"bootstrap_attempts": 200,
"bootstrap_draws": 200,
"cumulative_intensity_scale": 3.93674214,
"incidents": 100,
"observation_end_days": 100
},
"forecasts": [
{
"expected_incidents": 20.2368,
"expected_incidents_interval": {
"high": 26.4348,
"low": 15.1882
},
"homogeneous_rate_expected_incidents": 30,
"horizon_days": 30,
"probability_at_least_one_incident": 1
},
{
"expected_incidents": 56.9659,
"expected_incidents_interval": {
"high": 76.3508,
"low": 41.8621Truncated for display — the full payload is 60 lines.
How it works
Forecasting & survival — Estimate when something completes or fails, with censoring and unresolved work handled honestly rather than dropped.
- 1 Estimate long-run software reliability growth with a power-law nonhomogeneous Poisson process, bootstrap trend evidence, and future incident exposure.
- 2 Evaluate the method-specific diagnostics and gates returned by the function, then abstain unless the declared decision clears them under locally governed thresholds.
Before you trust it
Every tool in the catalog ships with the conditions under which its answer is meaningful — and the conditions under which it should abstain instead of guessing.
Assumptions & guardrails
- Training examples precede their outcomes, censoring and unresolved work are represented, and deployment populations remain comparable to validation cohorts.
- A predictive interval or risk estimate is not a deadline promise, causal explanation, or individual-performance judgment.
Minimum evidence
- systems: required and organization-defined
- horizons_after_observation_days: at least 1 rows/items
How to validate it
Validate on future periods or held-out aggregate units, compare with a simple baseline, and require stability across plausible metric definitions and decision thresholds.
Calibrating it to your org
Same for everyone
The mathematical kernel, validation rules, method version, and JSON output semantics are organization-independent; no tenant-trained coefficients or company benchmark is embedded in the function.
Specific to you
- incident time in days since a fixed window start
- forecast horizons
- incident inclusion/severity policy
Calibration workflow
- 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
- 2 Build a tenant-scoped historical cohort using only information available before each prediction or decision; preserve zero periods, censoring, assignment probabilities, and unresolved outcomes when the method requires them.
- 3 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
- 4 Validate on later time windows or held-out aggregate units at the deployment grain, against a simple baseline and the function-specific validation strategy.
- 5 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
- 6 Monitor realized outcomes, data drift, coverage, and decision regret; recalibrate at a governed cadence or after a detected regime/definition change, never merely because a stakeholder dislikes the result.
Call it from your AI
You don't wire up 388 tools in your MCP client. The GitRevio MCP server exposes 20 tools, three of which let an agent search the catalog, read a tool's schema, and run it — so the assistant finds this one on its own.
gitrevio_capabilities_search
{ "q": "estimate longrun software reliability growth with" }
→ finds "estimate_software_reliability_growth"
gitrevio_capability_describe
{ "capability_id": "estimate_software_reliability_growth" }
→ returns the input schema and agent guidance shown on this page
gitrevio_capability_run
{ "capability_id": "estimate_software_reliability_growth", "arguments": { ... } }
→ returns the result shown aboveWorks in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.