Playbooks

The Quick Start leaves you at the core loop: capture a request, send it to Repeater, change it, send it again. A playbook is what comes next — one workflow run start to finish, with a checkpoint at every step so you always know whether you're on track.

These are not reference pages. The Guide documents each tab in depth, one tool at a time; a playbook crosses tabs to finish a task the way an engagement actually does — scope before you capture, capture before you fuzz, confirm before you file. Read the Guide when you want to know what a tool is; work a playbook when you want to learn what to do with it.

Before you begin. Playbooks send real traffic. Point gori at a target you are authorized to test — your own app, a staging box, or a deliberately vulnerable practice target — and keep it in scope. Steps that need an exact result use a stable throwaway like example.com. Each step ends with a Checkpoint: what you should see before moving on.

Topics

Foundations — set the guardrails before you touch the target:

• Set up an engagement: a project, a scope, and a sandbox — why every active tool refuses to fire without them.
• Map the attack surface: build a sitemap, then let Discover find the paths you never clicked.

The manual loop — the core of hands-on testing:

• Intercept and modify in flight: hold a request, change it, forward it — then make the edit stick as a rule.
• Fuzz a parameter: mark a position, attach a wordlist, and read the results that stand out.
• Carry a session: extract a token once and replay authenticated across every request and every sweep.

Workbenches — focused tools, one job each:

• Decode and transform data: chain converters into a pipeline you can save and reuse.
• Attack a JWT: decode a token, tamper with its claims, and test whether the server checks the signature.
• Crack and forge session cookies: read a signed Flask/Rack/Django cookie, recover its secret, and mint your own.
• Grade token randomness: collect a few hundred session tokens and let the Sequencer score their predictability.
• Confirm blind vulnerabilities with OAST: plant an out-of-band payload and catch the callback that proves it fired.

Wrap up — turn findings into a report, or hand the project to an agent:

• Triage and report: file issues, prove a fix with the Comparer, and export a report a teammate can read.
• Run an AI co-pilot session: put an agent on the same project over MCP and watch every move it makes.

Next Steps

• Quick Start: the ten-minute path to the core loop, if you haven't run it yet
• Guide: in-depth reference for every tool a playbook touches
• Reference: every CLI subcommand, config key, and query-language filter

In This Section

• Attack a JWT
• Carry a session
• Confirm blind vulnerabilities with OAST
• Crack and forge session cookies
• Decode and transform data
• Fuzz a parameter
• Grade token randomness
• Intercept and modify in flight
• Map the attack surface
• Run an AI co-pilot session
• Set up an engagement
• Triage and report