Skip to content
Models
PLAN Free $0/mo PLAN Pro $20/mo PLAN Max $200/mo PLAN Teams $80/mo min + $40/seat Cognition FrontierCode (Oct 9) — score · cost per task Claude Opus 5.5 54.6% · $0.80/task Claude Fable 5 53.5% · $13.09/task Claude Opus 5 53.4% · $4.31/task GPT-6 Astra 53.3% · $4.59/task Claude Sonnet 5.5 52.1% · $1.59/task Claude Fable 5.1 50.9% · $3.28/task GPT-6.1 Sol 50.2% · $0.36/task SWE-2 50.0% · $1.18/task GPT-6 Sol 49.3% · $2.07/task Grok 4.6 48.0% · $2.88/task SWE-1.7 42.0% · $1.97/task SWE-1.6 9.4% · $0.49/task Devin CLI: v3000.11.92 Devin Cloud Updates Chat and Session Management Cognition adds Devin to Pylon support platform, offers $15,000 credits Cognition adds Gemini 4 Argon to Devin Desktop and CLI with 75% discount Devin CLI Update Devin Desktop: Continuity — 2026-10-12 Cognition launches FrontierCode 1.1 benchmark to assess AI‑generated p… Devin Cloud: Memory and dreaming · Session resume window is now 14 day… PLAN Free $0/mo PLAN Pro $20/mo PLAN Max $200/mo PLAN Teams $80/mo min + $40/seat Cognition FrontierCode (Oct 9) — score · cost per task Claude Opus 5.5 54.6% · $0.80/task Claude Fable 5 53.5% · $13.09/task Claude Opus 5 53.4% · $4.31/task GPT-6 Astra 53.3% · $4.59/task Claude Sonnet 5.5 52.1% · $1.59/task Claude Fable 5.1 50.9% · $3.28/task GPT-6.1 Sol 50.2% · $0.36/task SWE-2 50.0% · $1.18/task GPT-6 Sol 49.3% · $2.07/task Grok 4.6 48.0% · $2.88/task SWE-1.7 42.0% · $1.97/task SWE-1.6 9.4% · $0.49/task Devin CLI: v3000.11.92 Devin Cloud Updates Chat and Session Management Cognition adds Devin to Pylon support platform, offers $15,000 credits Cognition adds Gemini 4 Argon to Devin Desktop and CLI with 75% discount Devin CLI Update Devin Desktop: Continuity — 2026-10-12 Cognition launches FrontierCode 1.1 benchmark to assess AI‑generated p… Devin Cloud: Memory and dreaming · Session resume window is now 14 day…

Learn · Step 5 of 7

Knowledge, Playbooks, and Skills: teaching Devin your codebase

Three features that look interchangeable and are not. Knowledge is context, Playbooks are reusable prompts, Skills are procedures committed to the repo.

Once you are past your first few sessions, the bottleneck stops being how you phrase a task and becomes how much your agent already knows. Devin has three mechanisms for that, they are genuinely different, and picking the wrong one is the most common configuration mistake.

Knowledge — persistent context

Knowledge is a collection of tips, advice, and instructions Devin can reference across all sessions, recalled automatically when relevant.

Create it under the Knowledge tab in Settings & Library. Two fields are required:

  • Trigger Description — helps Devin recall the entry at the right moments
  • Content — a handful of sentences

Optionally add a Macro: a short identifier beginning with ! so you can pull the entry into a prompt directly.

Documented uses: code conformance practices, deployment workflows, PR naming conventions, testing workflows, and how to interact with proprietary internal tools. Devin will also suggest new Knowledge entries based on feedback you give it in chat — which means correcting it in conversation compounds instead of evaporating.

Reach for Knowledge when the answer is “this is how we do things here.”

Playbooks — reusable prompts

Playbooks are described as “easily shareable, reusable prompts for repeated tasks” — effectively a custom system prompt scoped to one recurring job.

The docs recommend a Playbook when the task will be reused across sessions, when you keep repeating the same reminders, or when the use case is relevant to other people on the team. Create one in the web app, or write a .devin.md file and drag it in.

There is an explicit boundary here: best practices and style guides belong in Knowledge, not Playbooks. Playbooks are for step-by-step procedures tied to a specific task.

Reach for a Playbook when the answer is “here is the procedure for this particular job.”

Skills — procedures as code

Skills are SKILL.md files committed to your repositories that teach Devin reusable procedures. They follow the open Agent Skills standard published at agentskills.io/specification.

The recommended location is .agents/skills/<category>/SKILL.md, and Devin discovers them automatically across connected repositories. They can be invoked explicitly by a user as /skill-name or autonomously by the agent, and notably a Skill can carry its own permission grants, restrictions, and even a different model.

Reach for a Skill when the procedure should be versioned, reviewed, and shared like code — because it is code, living in your repo and moving through pull requests.

Plugins — distributing Skills

For the CLI, a Plugin bundles Skills for installation from a GitHub repo, a git URL, or a local folder:

devin plugins install <source>

Plugins install at user level and are available across projects; their Skills are invoked as /:<skill> slash commands. The docs flag Plugins as beta, with behaviour and configuration subject to change — so build on them with that in mind.

Choosing between them

You want to encode…Use
Conventions, context, “how we do things”Knowledge
A repeatable procedure for one specific taskPlaybook
A procedure that should be versioned and reviewedSkill
A shareable bundle of SkillsPlugin

The failure mode to avoid is stuffing everything into prompts. If you are pasting the same paragraph into a third session, it belongs in one of the four rows above.

References

  1. Knowledge — docs.devin.ai
  2. Creating Playbooks — docs.devin.ai
  3. Skills — docs.devin.ai
  4. Plugins overview (CLI) — docs.devin.ai
  5. Agent Skills specification — agentskills.io

From Devin Central — subscribe for Devin news, features & tutorials. Devin is a trademark of Cognition.