Structured methodology for reliable results

Submit a task. Receive a structured deliverable.

Proven frameworks. Adaptive AI routing. Automated review and revision loops. The engine that turns prompts into deliverables you can trust.

Deliverable catalog

See it in action

Pick the deliverable you'd commission this week. See exactly what comes back.

Loading templates…

— or describe your own task —

20,000 FREE CREDITS

New task

No credit card needed.

Verifiability

You get a good answer. You still have to vouch for it.

Here is one task we ran, what each step of it concluded, and what each side leaves behind.

“Run a competitive analysis of the project management software market to inform our product strategy. Assess how Asana, Monday.com, ClickUp, and Linear are positioned — including their AI features — along with their strengths and where they're vulnerable, then recommend where we should focus to differentiate.”

A chat answer

What you get1 replyFluent, well structured, and usually right. That is not the part in question.

What you can openNothingNo plan, no record of what each step produced, no source list, and no cost for the work.

What it costs youYour timeYou reconstruct the reasoning and re-check the claims yourself before you can put the work in front of anyone.

What you hand your clientYour wordYou defend the conclusion in the room, without being able to show how it was reached or what it rests on.

A Rigor run

ClickUp offers granular customization with complexity trade-offs, Asana provides structured simplicity with scaling limitations…3m 25s1 · Research

All four have shipped AI capabilities, but these remain primarily tactical … rather than strategic.1m 40s2 · Gap analysis

Recommendation: YES (with conditions) … The primary risk is execution speed, not market existence.1m 8s3 · Strategic evaluation

…a shared blind spot: none deliver portfolio-level AI intelligence…5m 10s4 · Synthesis

The synthesis claims “no incumbent has built” portfolio AI, but at minimum Epicflow appears to be attempting this.1m 17s5 · Review protocol

What you hand your client20 sourcesThe deliverable itself, every source under it, and all 5 steps that produced it — open beside you in the room.

Every line in the right-hand column is quoted from run wr_a3bb84ccc118b52845e2ec4c, a Competitive scan we ran on 2026-05-31. Free accounts see the plan, every step with its credits and runtime, and the deliverable with its sources. Full step output and the method behind each step are on Pro and above.

Users

Built for how you work

One layer. Three entry points.

For business professionals

Structured deliverables — competitive scans, gap assessments, SWOT analyses, solution designs.
Quality gates on every step. Act on the result, not raw AI output.
Project folders and export. Organize by client, share with your team.

Try it free →

For developers

Code reviews, solution designs, and implementation prompts through proven methodology.
Connect GitHub for codebase-aware analysis. Step-by-step mode for full control.
API-first with MCP server + REST. Everything the dashboard does, the API does too.

View API docs →Try the dashboard →

For AI agents

Structured methodology for every task. Same rigorous standard, every time.
Quality gates and adaptive routing — errors caught before they propagate.
Single API, MCP server + REST credit-based billing. No orchestration code required.

Inside the engine

How a prompt becomes a deliverable

A chat completion is one prompt. Rigor is a stack. Watch it run.

Classify & scope55+ task types · 3-tier classification · automatic rigor level

Compose methodology15+ proven frameworks · sequenced per task type · not one-size-fits-all

Route intelligently10+ LLM providers · best model per step · optimizes cost, quality, and speed

Execute & gateevery step quality-checked · refine and re-execute · model escalation on failure

Adapt & learnrouting improves with every task · the engine gets smarter over time

Integrations

Plug into your stack

Rigor reads your codebase, docs, and tickets automatically — no copy-pasting.

01ReadContext from your stack
→
02RunMethodology with full project context
→
03WriteResults back where your team works

↓GitHubrepos · issues · PRs

↓ ↑Claude Codecli sessions · skills

↓Notionpages · databases

↓Linearissues · cycles · projects

↓Jiraepics · stories · sprints

↓Confluencespaces · pages

↓Figmafiles · frames · specs

↓ ↑Slackchannels · threads · DMs

Workspace-scoped. Connections are isolated per workspace — your project context never crosses boundaries. ↓ ↑ read & write · ↓ read-only

Direct execution

The cheapest model that holds your schema.

Declare the shape you want back. Rigor picks the cheapest model that measurably holds it.

Classifytag · quick_classification · 1 label or many, from a closed set you define

Scorescore · relevance and quality judgments, on your scale

Rankrerank · a permutation of your own candidate ids, validated before return

Extract & parseextract_entities · parse_query · typed records out of free text

Route on evidenceevery answer scored against your contract · cheaper models promoted on the evidence

Direct execution in the API docs →Declare an output_contract →

Pricing

One methodology layer. Pay for what you run.

Start with 20,000 free credits. Upgrade when you need more.

Free$0/mo

Propopular$15/mo

Team$79/mo

Enterprise$399/mo

Members

Max workspaces

Unlimited

API keys / workspace

Unlimited

MCP connections

10/workspace

Unlimited

History

7 days

90 days

Unlimited

Concurrent workflows

Unlimited

API rate limit

120/min · 1,000/day

500/min · 100,000/day

Interactive mode (Mode C)

—

Webhooks

—

Step outputs visible

—

Show model names

—

LLM policy

—

Data residency

—

Workspace budgets

—

Audit log export

—

Usage pricing

Pay for what you run

Your subscription covers the platform — dashboard, MCP server, integrations, tier-level features. Credits cover the AI work itself. Every workflow consumes credits based on complexity and model routing; you see the running cost in real time.

Same per-credit price on every tier · credits never expire · See the full pricing →

All plans include the full engine, dashboard, and public MCP server. Tasks consume credits at the same per-credit price on every tier.

Encrypted in transit and at rest · Row-level data isolation · No training on your data · GDPR-compliant deletion · LLM policy controls on Team and Enterprise · Data residency on Enterprise.

Developer API

Call Rigor from your code, your agent, your editor.

One REST endpoint. One MCP server. cURL, raw HTTP, or drop into your agent.

cURLMCP

# Plan a task (free)
curl -X POST https://plith.ai/api/rigor/plan \
  -H "x-api-key: $PLITH_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "task_description": "Competitive scan of NovaTech — top three rivals"
  }'

# Execute with SSE streaming
curl -X POST https://plith.ai/api/rigor/execute \
  -H "x-api-key: $PLITH_KEY" \
  -H "Content-Type: application/json" \
  -H "Accept: text/event-stream" \
  -d '{
    "task_description": "Competitive scan of NovaTech — top three rivals",
    "preferences": { "rigor_level": "standard" }
  }'

# → plan in <5s · deliverable streamed via SSE
# → quality score + cost breakdown on every step

As a former strategy consultant I noticed that deliverables created by AI tools always proved inconsistent and required manual interventions.
So I built Rigor as an intelligent router to apply best practice sequences of proven frameworks — for reliable, high-quality deliverables.

Lukas — Founder, Plith

Try it now

See the dashboard you'll be working in.

Library, workspaces, integrations, billing — wired with sample tasks.

Open the dashboard →View a demo

Sample data · read-only · no signup required