The product tour

One AI that plansyour software,
from PRD to production, on one graph.

Thirteen sections, four modules, sixty-plus shipping capabilities. No salespeak. Every visual on this page is a render of the actual product, on the actual data shapes, with the actual numbers.

End-to-end AI workflow

One PRD in. A shipped release out.

Five live surfaces: wizard, backlog, architecture, tests, release. Scroll through the whole loop in 90 seconds.

  1. PRD
  2. Backlog
  3. Architecture
  4. Verify
  5. Release

Drop the problem in.

6-step wizard fills the brief: problem, audience, success metrics, scope, non-goals, open questions.

stride.page/projects/aurora/prd-generate
Problem drafted

Enterprise admins cannot enforce SSO. Today they rely on per-user email/password, which fails 17% of SOC 2 audits.

Audience drafted

Workspace owners on the Business + Enterprise tier. Identity admins. Compliance leads.

Success metrics drafted

• Time-to-SSO < 10 min • SCIM JIT match rate ≥ 98% • Zero password-resets/30d on enabled workspaces

Scope

IdP metadata XML upload · SP-initiated + IdP-initiated · SCIM 2.0 user + group sync · Audit log emission

Non-goals

LDAP. Just-in-time access requests. Per-IdP claim transformation.

Open questions

• What's the right behaviour for existing API keys at SSO enable? • Should owners require step-up for billing changes?

stride.page⌘K · ask AI
Plan

Plan, backlog, and ship, with AI in the loop.

Sprint planning, AI Issue Writer, dependency tracking, release notes: every artifact on one graph the agents can traverse.

Kanban + sprints + releases + backlogAI Issue Writer with confidence + toneDefect-prediction priority on every storyAI release notes with version historyStory dependencies (graph)Threaded comments + @mentionsLive presence + SSE activity (20s ping)Velocity + burndown analytics
Design

Architecture that thinks alongside you.

AI Solutions scored on five axes, ADRs with version history, real PR drafts from architectural findings, and six diagram types, sequence to deployment.

AI Solutions (3-5 scored options)AI Architectural Review w/ PR draftADRs with full version historyArchitecture Fitness (6 functions, targets)Tech Radar (Adopt/Trial/Assess/Hold)C4 + Sequence + Data flow + DeploymentDependency matrixReal GitHub branch + PR creation
Optimize

Find the bottleneck. Replace the legacy. Calculate the ROI.

BPMN process mining with bottleneck heatmaps, automation discovery with payback math, and Legacy Intelligence: AI that reads decades-old code and writes the modernization plan.

BPMN process miningBottleneck detection w/ heatmapAutomation kanban + ROI mathWorkload + capacity donutLegacy Intelligence · 8-tab dashboardCOBOL · PHP · Java demo systemsModernization roadmap w/ $ paybackStrategic Q&A over legacy code
Verify

Tests written by AI. Quality measured continuously.

AI streams Gherkin scenarios from your issues. Traceability shows every issue's coverage in one heatmap. Defect severity, MTTR, quality gates ride alongside.

AI test generation w/ Gherkin SSE streamTest pyramid · 1,842 unit / 184 int / 38 e2eTraceability matrix (FULL/PARTIAL/NONE/FAILING)Defect analytics by severityMTTR sparklineQuality gates pipelineCoverage gap detectorFlaky test aggregation
Agents · MCP · Cost-cap

Not a chatbot. A fleet of agents, on a real server, on a leash.

Stride is an MCP server with a tool for everything you can do in the UI. The /agent console runs them, the cost-cap pins them to a budget. Every AI call atomic at the database.

Stride as MCP server live
# ~/.config/mcp/clients.json
{
  "stride": {
    "url": "https://www.stride.page/api/mcp/stride",
    "headers": { "Authorization": "Bearer $STRIDE_TOKEN" }
  }
}

# claude: list available tools
$ claude mcp tools stride
# → a tool for every workspace object (Plan-MCP P0–P3)

A tool for every object · scroll to see more

· list_stories· create_story· update_story· create_sprint· update_sprint· transition_sprint· get_sprint_metrics· create_defect· update_defect· create_story_relation· list_story_relations· delete_story_relation· add_comment· list_comments· create_release· list_releases· add_attachment· list_attachments· list_members· invite_member· update_member_role· list_labels· create_label· apply_label· list_epics· create_epic· bulk_update_stories· search_workspace· get_audit_log· create_webhook
Agent consolestride.page/agent

Agents · 7 registered

PRD Writer142
Issue Author481
Architectural Reviewer38
Test Author219
Release Notes Author24
Process Optimizer16

Run trace · Issue Author

1,205 tokens · 2.4s
  1. list_stories18ms122 tk
  2. get_story9ms0 tk
  3. search_workspace64ms388 tk
  4. create_story41ms504 tk
  5. apply_label12ms0 tk
  6. add_comment31ms191 tk
Cost cap advisory-lock · 402-safe
Per-run cap1,204 / 2,500 tokens
// lib/ai/ai-metering.ts
await db.$transaction(async (tx) => {
  await tx.$queryRaw`SELECT pg_advisory_xact_lock(${wid})`;
  const reserved = await reserve(tx, wid, est);
  if (reserved > budget) throw new BudgetExceeded402();
  // … invoke model …
  await reconcile(tx, wid, actualCost);
});

Atomic Postgres advisory lock per workspace. No double-spend, ever. A budget overrun returns HTTP 402 with a Retry-After hint and a link to upgrade. Agents handle it gracefully.

One server, two runtimes. claude and codex point at the same MCP endpoint; your team picks the editor, Stride stays the source of truth. live telemetry · 12 active runs across workspace

Stride MCP server, full workspace coverageClaude Code + Codex clientsAgent console · Chat / Operate / HistoryPer-run token + cost capPostgres advisory-lock metering402 graceful failure on capRun trace with per-tool spendTool surface = Story · Sprint · Defect · Comment · Label · Member
One graph

One issue. Everything it touches, one hover away.

Plan, Design, Optimize, Verify, and the agent fleet: five clusters, one Prisma-backed graph. Hover any node to see what connects to it. That's why one AI can read the whole story in a single prompt.

Cross-module knowledge graph, hover to traverseIssue · STR-247Sprint 14Release v3.4EpicBacklogRoadmapADR-014C4 DiagramAI SolutionFitnessTech radarSupport flowAutomationCOBOL coreBottleneckMod. roadmapTC-901GherkinDefectPyramidTraceabilityPRD agentTest agentMCP server
PlanDesignOptimizeVerifyAgents
47Plan

acceptance criteria across the sprint

5/6Design

fitness gates passing · 14 ADRs versioned

$2.1MOptimize

annual savings on the COBOL roadmap

1,842Verify

unit tests · 99.6% overall pass

51Agents

MCP tools exposed to Claude + Codex

Differentiators

Five things you can't do in any other tool.

#1

One graph, not four tools

Every issue, ADR, diagram, process, test, automation, and legacy component lives on the same graph. The AI traverses it in a single prompt.

One-graph mini-illustration
#2

End-to-end AI, not a chatbot bolted on

PRD wizard → epics → issues → architecture → tests → release notes. Each stage feeds the next; AI is the connective tissue, not a sidebar.

PRD
Backlog
Arch
Tests
Release
#3

Legacy intelligence nobody else ships

Point Stride at a 38-year-old COBOL codebase. Get an 8-tab dashboard, extracted business rules, a derived data model, and a modernization roadmap with $ payback.

PROGRAM-ID. ACCT-XFER.
MOVE 100 TO X.
transferFunds(input)
  return tx.run()
#4

Process mining with ROI math

BPMN heatmap finds your bottleneck. Automation kanban quantifies the payback. Inline calculators show inputs × time × rate, with confidence bands.

$462kautomation savings / yr
#5

Cost-cap on every AI call, atomic at the DB

Postgres advisory lock per workspace. 402 graceful failure when the budget is hit. No double-spend, ever. Your CFO sleeps. Your agents stay on a leash.

advisory-lock · 402-safe
Ecosystem

Plugs into your stack. Earns your audit team's trust.

Six shipping integrations today. Honest list, no fake logos. RBAC, IP allow-list and a full audit log out of the box, with SOC 2 Type I in progress. Developer surface deep enough to script anything we ship.

Integrations · 6 shipped

GitHubLive
Source control

PR ↔ Issue linking, status on merge, GitHub App install.

SlackLive
Communication

@mentions, issue status, sprint events. Per-category toggles.

Jira CloudLive
Project management

Two-way sync of epics, issues, sprints, comments. OAuth.

Claude CodeLive
AI coding agent

Your backlog in the CLI/IDE. Agents read + update via MCP.

OpenAI CodexLive
AI coding agent

Same MCP server, native Codex support. One key, two runtimes.

Webhooks · REST · GraphQLLive
Build your own

HMAC-signed events + REST + GraphQL. No waiting on us.

Security · RBAC

  • SOC 2 Type I

    in progress, target 2026

  • Audit log

    every state change, queryable + exportable

  • RBAC + 2FA

    owner / admin / member / guest, TOTP enforced

  • IP allow-list

    per-workspace CIDRs, breaks fail-closed

  • Encryption

    AES-256 at rest, TLS 1.3 in transit

Developer surface

  • MCP server

    /api/mcp/stride · a tool for every object

  • REST + GraphQL

    full coverage, OpenAPI 3.1 spec

  • Webhooks

    HMAC-signed, replay-safe, per-category

  • API keys · scopes

    tokens with explicit grants, rotatable

  • Sandbox

    scoped sandbox keys, every API

Ready to plug your team into the graph?

Stride starts at $9/seat/month, paid from day 1. AI credits included. Cancel any time.

Get started · $9/moSee pricing
Keep exploringPricingBlogStride vs LinearStride vs JiraStride vs ClickUp
← back to home back to top