Learn one task at a time

1. Run work across agents

   SpecWeave CLI

   Plan an increment, prove each task, hand off between tools and get a review from another model.

   1. Plan, prove and hand off a task with the CLI

      1:00

   2. Install SpecWeave and plan your first increment

      1:36

   3. Claim, prove, verify and complete a task

      1:30

   4. Hand off from Claude Code to Codex at a usage limit

      1:18

   5. Review with the other model before you ship

      1:33

2. Run SpecWithCode

   SpecWithCode

   Understand the setup, run a project through its coordinator, then add plans, memory and a second provider.

   1. See how SpecWithCode runs your agents

      2:54

   2. Run a project with a coordinator and workers

      1:29

   3. Plan, verify and schedule project work

      1:35

   4. Keep project memory and read honest usage

      1:06

   5. Continue a thread on Codex at a usage limit

      1:01

   5 more guides are planned for this path.

17 guides

• Plan, prove and hand off a task with the CLI

  You see the whole loop once: the plan, the claim, the test evidence and the handoff all travel with the repository.

  SpecWeave CLIAny agentNarratedRecorded on SpecWeave 3.0.6, added 12 October 2026

  1:00

• Install SpecWeave and plan your first increment

  You end with an increment whose spec.md lists numbered acceptance criteria and tasks with their files and test, and specweave pickup tells you which task to claim first.

  SpecWeave CLIAny agentNarratedRecorded on SpecWeave 3.1.0, 18 October 2026, added 18 October 2026

  1:36

• Claim, prove, verify and complete a task

  You end with a completed increment that has its test evidence and a passing verify report.

  SpecWeave CLIAny agentNarratedRecorded on SpecWeave 3.1.0, 18 October 2026, added 18 October 2026

  1:30

• Hand off from Claude Code to Codex at a usage limit

  You end with Codex running the next task from the ledger, and the ledger shows who handed off, who picked up and who finished.

  SpecWeave CLIClaude Code and CodexNarratedRecorded on SpecWeave 3.1.0, 18 October 2026, added 18 October 2026

  1:18

• Review with the other model before you ship

  You end with a review report from a model of another family and a clean second run.

  SpecWeave CLIClaude Code and CodexNarratedRecorded on SpecWeave 3.1.0 and codex-cli 0.162.1, 18 October 2026, added 18 October 2026

  1:33

• See how SpecWithCode runs your agents

  You know how machines, subscriptions, usage limits, the coordinator and its threads fit together before you install.

  SpecWithCodeClaude Code and CodexNarratedRecorded on SpecWithCode 1.0.5, added 13 October 2026

  2:54

• Run a project with a coordinator and workers

  You can tell the coordinator and worker models apart, follow the routing history, resolve and reopen a thread, and read token usage by model.

  SpecWithCodeClaude Code and CodexCaptions, no soundOlder interface, added 9 October 2026

  1:29

• Plan, verify and schedule project work

  You add a plan with one acceptance criterion, see a failed verification stay failed, record a passing attempt, and keep routines scoped to their project.

  SpecWithCodeAny agentCaptions, no soundOlder interface, added 9 October 2026

  1:35

• Keep project memory and read honest usage

  You keep a project goal, instructions and decisions in small memory files and read usage that comes only from recorded provider turns.

  SpecWithCodeAny agentCaptions, no soundOlder interface, added 9 October 2026

  1:06

• Continue a thread on Codex at a usage limit

  You see a coordinator split work between Codex and Claude, and a Claude thread at its limit continue on Codex with its context.

  SpecWithCodeClaude Code and CodexCaptions, no soundOlder interface, added 10 October 2026

  1:01

• Set up a personal project with folders and tools

  You create a personal project, add a folder and a repository, and grant a local MCP command exactly the tools it may use.

  SpecWithCodeAny agentCaptions, no soundOlder interface, added 9 October 2026

  1:21

• Check what an agent actually did

  You can find the recorded tool call, its exit code and the file change behind an agent’s answer.

  SpecWithCodeAny agentCaptions, no soundOlder interface, added 8 October 2026

  0:43

• Run your first task with Claude Code

  You end with Claude Code having planned, built and tested one change in your repository, with the ledger line that records who finished it.

  SpecWithCodeClaude CodeNarratedOlder interface, added 11 October 2026

  1:06

• Run your first task with Codex

  You end with Codex having planned, built and tested one change in your repository, with the ledger line that records who finished it.

  SpecWithCodeCodexNarratedOlder interface, added 11 October 2026

  1:12

• Run your first task with OpenCode

  You end with OpenCode having planned, built and tested one change in your repository, with the ledger line that records who finished it.

  SpecWithCodeOpenCodeNarratedOlder interface, added 11 October 2026

  1:10

• Run your first task with Antigravity and Gemini

  You end with Antigravity having planned, built and tested one change in your repository, with the ledger line that records who finished it.

  SpecWithCodeAntigravityNarratedOlder interface, added 11 October 2026

  1:14

• Run your first task with Grok

  You end with Grok having planned, built and tested one change in your repository, with the ledger line that records who finished it.

  SpecWithCodeGrokNarratedOlder interface, added 11 October 2026

  1:23

Guides for finding, testing and publishing agent skills live on verified-skill.com. Verified Skill tutorials

• A real recordingFootage of the running app or real command output, with sample projects. Each guide says which version it shows and when it was added; recordings of the current release come first.
• Captions, chapters and a transcriptA WebVTT caption track, chapters you can jump to with the keyboard, and a transcript built from the captions.
• Hosted on this siteVideos and posters load from spec-weave.com, with no third-party player or tracking.