Validated source of truthfor AI-built products

The spec your AI agent builds against — and keeps true.

VasperaPM reverse-engineers a spec from your existing code, validates it as canonical, and continuously verifies the code never drifts. Your agent reads one validated set instead of re-deriving it every session.

vaspera-pm-agent
verify_docs_code--against canonical

Loading canonical spec set...

Verifying 230 requirements against code...

Tracing each requirement to implementation...

Scoring drift and confidence...

1 drifted · 74% confidence · canonical set ready

→ 229/230 requirements verified
→ REQ-114 drifted from spec
→ /reports/drift.md
→ canonical set your agent reads via get_canonical_backlog

33
verification & PM tools
4
drift categories tracked
3
surfaces: CLI, MCP, CI
60s
to your first drift report

The validated source of truth your AI keeps building against

Verification at the core, with the PM tooling to act on what it finds, from drift detection to backlog generation.

Code-First Intelligence

No PRD? Scan your prototype repo to auto-infer product goals, user flows, and API contracts.

Validated source of truth

Specs carry a lifecycle — draft → proposed → validated. Validated specs are canonical: the set your AI agent builds against instead of re-deriving them.

Full Backlog Explosion

Generate epics → features → stories → Gherkin acceptance criteria with MoSCoW priorities and effort estimates.

Architecture Diagrams

Auto-create Mermaid/PlantUML diagrams: system architecture, data flows, and sequence diagrams for key epics.

Dev Handoff Ready

Create branches, commit artifacts, open PRs, and export to Jira/Linear automatically.

Gap Validation

Compare inferred specs vs code — highlights missing error handling, edge cases, and GDPR compliance.

Enterprise Security

CSP protection, encrypted tokens, rate limiting, Sentry monitoring, and SOC 2-ready audit logs.

Continuous drift verification

Every requirement is re-checked against live code. When an agent's change drifts from the spec, you see exactly which requirement broke.

Verify any repo from the browser

Our server-side runner checks public and private repos and reports per-requirement drift — no local clone.

Code-first

Code-to-Specs Reverse Engineering

Bidirectional by design. Point VasperaPM at a messy prototype and it reverse-engineers the spec it implements — then drives building spec→code and verifies the code never drifts from it. Other spec tools only go forward, from a blank page.

Handles code-first prototypes

infer_prd_from_code

Scans your entire repo, analyzes structure and API calls, infers product goals and gaps — generates a master PRD automatically.

Visualizes implicit flows

reverse_engineer_user_flows

Traces routes, components, and events to map user journeys. Outputs Mermaid flowcharts and user story outlines.

Critical for backend planning

import_verify_spec

Extracts fetch/axios calls and mock data to generate OpenAPI specs and required backend endpoint schemas.

Prototype to production

run_pipeline

One-click pipeline: full explosion → epics → stories → AC → diagrams → implementation plans.

Ensures commercial readiness

validate_implementation

Compares the spec vs code — highlights missing error handling, edge cases, accessibility, and compliance gaps.

Maximum flexibility

Dual-Path Pipeline

Works with docs-first OR code-first approaches. Auto-detects what you have and fills in what's missing.

Code-First Pipeline
# No docs? Start from code
infer_prd_from_code(
  root_path="./my-prototype",
  focus="commercialization"
)

# Chain into full pipeline
validate_implementation(
  focus_areas=["backend", "security", "monetization"]
)

# Validate before shipping
verify_docs_code()

# Output: Full /backlog/ + /diagrams/ + handoff package

From chaos to clarity in minutes

Two paths to production-ready backlogs. Start with docs or start with code — VasperaPM handles both.

01

Connect Your Docs

Point VasperaPM at your /docs folder. It scans PRDs, specs, notes, and prototype files automatically.

synthesize_master_prd(
  folder_path="/docs",
  include_patterns=["*.md", "*.pdf"]
)
02

Synthesize & Order

AI analyzes content, extracts timelines, resolves conflicts, and creates a unified master PRD.

synthesize_master_prd(
  ordered_docs_path="./ordered_docs",
  resolve_contradictions=True
)
03

Explode the Backlog

Generate complete backlogs with epics, stories, and acceptance criteria focused on your priorities.

explode_backlog(
  prd_path="./PRD_master.md",
  focus_areas=["backend", "security"]
)
04

Handoff to Dev

Create branches, commit artifacts, open PRs, and export directly to Jira or Linear.

handoff_package(
  package_name="v1.0-release",
  create_pr=True
)

Simple, transparent pricing

Start free, scale as you grow. No hidden fees.

Free

$0forever

100 tool calls/month to explore

  • 100 tool calls per month
  • All 33 MCP tools
  • Doc-code drift detection
  • Community support
Get Started

Starter

$9/month

For solo PMs and small teams

  • 500 tool calls per month
  • All 33 MCP tools
  • Jira/Linear/GitHub export
  • Email support
Start Free Trial
Most Popular

Pro

$29/month

For teams shipping products fast

  • 2,500 tool calls per month
  • All 33 MCP tools
  • Jira/Linear/GitHub/ADO export
  • Priority support
  • Historical drift tracking
Start Free Trial

Enterprise

Custompricing

For organizations at scale

  • Unlimited tool calls
  • Everything in Pro
  • SSO/SAML
  • Sentry monitoring & audit logs
  • Dedicated support
  • Self-hosted option
Contact Sales

Give your AI one spec to build against.

Point VasperaPM at your repo and get a validated spec your AI can build against — with drift caught the moment it appears. Start free, no credit card.