Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 8 additions & 0 deletions CHANGELOG.md
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,14 @@ All notable changes to Agent OS will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).

## [2.0.3] - 2025-10-10

- Updated instructions and default standards to reduce excessive tests writing and test running during feature development to improve speed and token useage.
- For Claude Code users:
- Replaced hard-coding of 'opus' model setting on agents with 'inherit' so that it inherits whichever model your Claude Code is currently using.
- Updated create-role script to add the "Inherit" option when creating new agents.
- Clarified next command to run when in single-agent mode.

## [2.0.2] - 2025-10-09

- Clarified /create-spec command so that task list creation doesn't begin until spec.md has been written.
Expand Down
2 changes: 1 addition & 1 deletion config.yml
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
version: 2.0.0
version: 2.0.3
base_install: true

# ================================================
Expand Down
2 changes: 1 addition & 1 deletion profiles/default/agents/implementation-verifier.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ name: implementation-verifier
description: Verify the end-to-end implementation of a spec
tools: Write, Read, Bash, WebFetch, Playwright
color: green
model: opus
model: inherit
---

You are a product spec verifier responsible for verifying the end-to-end implementation of a spec, updating the product roadmap (if necessary), and producing a final verification report.
Expand Down
2 changes: 1 addition & 1 deletion profiles/default/agents/product-planner.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ name: product-planner
description: Create product documentation including mission, and roadmap
tools: Write, Read, Bash, WebFetch
color: cyan
model: opus
model: inherit
---

You are a product planning specialist. Your role is to create comprehensive product documentation including mission, and development roadmap.
Expand Down
2 changes: 1 addition & 1 deletion profiles/default/agents/specification/spec-researcher.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ name: spec-researcher
description: Gather detailed requirements through targeted questions and visual analysis
tools: Write, Read, Bash, WebFetch
color: blue
model: opus
model: inherit
---

You are a software product requirements research specialist. Your role is to gather comprehensive requirements through targeted questions and visual analysis.
Expand Down
2 changes: 1 addition & 1 deletion profiles/default/agents/specification/spec-writer.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ name: spec-writer
description: Create a detailed specification document for development
tools: Write, Read, Bash, WebFetch
color: purple
model: opus
model: inherit
---

You are a software product specifications writer. Your role is to create a detailed specification document for development.
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ name: task-list-creator
description: Create a detailed and strategic tasks list for development of a spec
tools: Write, Read, Bash, WebFetch
color: orange
model: opus
model: inherit
---

You are a software product tasks list writer and planner. Your role is to create a detailed tasks list with strategic groupings and orderings of tasks for the development of a spec.
Expand Down
6 changes: 3 additions & 3 deletions profiles/default/agents/templates/verifier.md
Original file line number Diff line number Diff line change
Expand Up @@ -48,11 +48,11 @@ Read the following files to understand the user's standards and preferences so t

{{verifier_standards}}

### Step 4: Run the tests that were written for the tasks under your verification purview
### Step 4: Run ONLY the tests that were written by the implementer of the tasks under your verification purview

IF the tasks under your verification purview involved writing of tests, then run ONLY those specific tests and note how many are passing and failing.
IF the implementer of the tasks under your verification purview wrote tests that cover this implementation, then run ONLY those specific tests and note how many are passing and failing. Do NOT run the entire app's tests suite.

If any are failing then note the failures, but DO NOT try to implement fixes.
If any tests are failing then note the failures, but DO NOT try to implement fixes.

### Step 5: (if applicable) view and screenshot the implemented features in a browser

Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,9 @@ The tasks list has created at `agent-os/specs/[this-spec]/tasks.md`.

Review it closely to make sure it all looks good.

Next step: Run the command, 3-verify-spec.md to closely verify your spec and tasks list for accuracy and alignment. Or you can skip straight to running the implement-spec.md command if you're ready.
Next step: Run the command, `3-verify-spec.md`.

Or if want, you can skip straight to running the `implement-spec.md` command.
```

## User Standards & Preferences Compliance
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ Your spec verification report is ready at `agent-os/specs/[this-spec]/verificati

Review it closely to make sure it all looks good.

Next step: Run the command, implement-spec.md to generate prompts for implementation.
Next step: Run the command, `implement-spec.md` to generate prompts for implementation.
```

## User Standards & Preferences Compliance
Expand Down
10 changes: 10 additions & 0 deletions profiles/default/commands/new-spec/single-agent/1-new-spec.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,3 +3,13 @@ This begins a multi-step process for planning a new spec for our next product in
The FIRST STEP is to initialize the spec by following these instructions:

{{workflows/specification/initialize-spec}}

## Display confirmation and next step

Once you've initialized the spec folder, output the following message (replace `[this-spec]` with the folder name for this spec)

```
✅ I have initialized the spec folder at `agent-os/specs/[this-spec]`.

Next step: Run the command, 2-research-spec.md
```
10 changes: 10 additions & 0 deletions profiles/default/commands/new-spec/single-agent/2-research-spec.md
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,16 @@ Follow these instructions for researching this spec's requirements:

{{workflows/specification/research-spec}}

## Display confirmation and next step

Once you've completed your research and documented it, output the following message:

```
✅ I have documented this spec's research and requirements in `agent-os/specs/[this-spec]/planning`.

Next step: Run the command, `1-create-spec.md`.
```

After all steps complete, inform the user:

"Spec initialized successfully!
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -5,3 +5,19 @@ The FIRST STEP is to confirm the product details by following these instructions
{{workflows/planning/gather-product-info}}

Then WAIT for me to give you specific instructions on how to use the information you've gathered to create the mission and roadmap.

## Display confirmation and next step

Once you've gathered all of the necessary information, output the following message:

```
I have all the info I need to help you plan this product.

Next step: Run the command, `2-create-mission.md`
```

## User Standards & Preferences Compliance

When planning the product's tech stack, mission statement and roadmap, use the user's standards and preferences for context and baseline assumptions, as documented in these files:

{{standards/global/*}}
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,18 @@ Now that you've gathered information about this product, use that info to create

{{workflows/planning/create-product-mission}}

## Display confirmation and next step

Once you've created mission.md, output the following message:

```
✅ I have documented the product mission at `agent-os/product/mission.md`.
Review it to ensure it matches your vision and strategic goals for this product.
Next step: Run the command, `3-create-roadmap.md`
```

## User Standards & Preferences Compliance

IMPORTANT: Ensure the product mission is ALIGNED and DOES NOT CONFLICT with the user's preferences and standards as detailed in the following files:
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,18 @@ Now that you've created this product's mission.md, use that to guide your creati

{{workflows/planning/create-product-roadmap}}

## Display confirmation and next step

Once you've created roadmap.md, output the following message:

```
✅ I have documented the product roadmap at `agent-os/product/roadmap.md`.

Review it to ensure it aligns with how you see this product roadmap going forward.

Next step: Run the command, `4-create-tech-stack.md`
```

## User Standards & Preferences Compliance

IMPORTANT: Ensure the product roadmap is ALIGNED and DOES NOT CONFLICT with the user's preferences and standards as detailed in the following files:
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,18 @@ The final part of our product planning process is to document this product's tec

{{workflows/planning/create-product-tech-stack}}

## Display confirmation and next step

Once you've created tech-stack.md, output the following message:

```
✅ I have documented the product's tech stack at `agent-os/product/tech-stack.md`.
Review it to ensure all of the tech stack details are correct for this product.
You're ready to start planning a feature spec! You can do so by running the command, `1-new-spec.md`.
```

## User Standards & Preferences Compliance

The user may provide information regarding their tech stack, which should take precidence when documenting the product's tech stack. To fill in any gaps, find the user's usual tech stack information as documented in any of these files:
Expand Down
8 changes: 4 additions & 4 deletions profiles/default/roles/implementers.yml
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ implementers:
description: Handles migrations, models, schemas, database queries
your_role: You are a database engineer. Your role is to implement database migrations, models, schemas, and database queries.
tools: Write, Read, Bash, WebFetch
model: opus
model: inherit
color: orange
areas_of_responsibility:
- Create database migrations
Expand All @@ -29,7 +29,7 @@ implementers:
description: Handles API endpoints, controllers, business logic, request/response handling
your_role: You are an API engineer. Your role is to implement API endpoints, controllers, business logic, and handle request/response processing.
tools: Write, Read, Bash, WebFetch
model: opus
model: inherit
color: blue
areas_of_responsibility:
- Create API endpoints
Expand All @@ -55,7 +55,7 @@ implementers:
description: Handles UI components, views, layouts, styling, responsive design
your_role: You are a UI designer. Your role is to implement UI components, views, layouts, styling, and ensure responsive design.
tools: Write, Read, Bash, WebFetch, Playwright
model: opus
model: inherit
color: purple
areas_of_responsibility:
- Create UI components
Expand All @@ -81,7 +81,7 @@ implementers:
description: Handles test files, test coverage, test fixtures
your_role: You are a testing engineer. Your role is to write comprehensive tests for features that have been implemented by other engineers.
tools: Write, Read, Bash, WebFetch
model: opus
model: inherit
color: green
areas_of_responsibility:
- Write unit tests
Expand Down
7 changes: 0 additions & 7 deletions profiles/default/standards/testing/coverage.md

This file was deleted.

9 changes: 9 additions & 0 deletions profiles/default/standards/testing/test-writing.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,9 @@
## Test coverage best practices

- **Write Minimal Tests During Development**: Do NOT write tests for every change or intermediate step. Focus on completing the feature implementation first, then add strategic tests only at logical completion points
- **Test Only Core User Flows**: Write tests exclusively for critical paths and primary user workflows. Skip writing tests for non-critical utilities and secondary workflows until if/when you're instructed to do so.
- **Defer Edge Case Testing**: Do NOT test edge cases, error states, or validation logic unless they are business-critical. These can be addressed in dedicated testing phases, not during feature development.
- **Test Behavior, Not Implementation**: Focus tests on what the code does, not how it does it, to reduce brittleness
- **Clear Test Names**: Use descriptive names that explain what's being tested and the expected outcome
- **Mock External Dependencies**: Isolate units by mocking databases, APIs, file systems, and other external services
- **Fast Execution**: Keep unit tests fast (milliseconds) so developers run them frequently during development
10 changes: 0 additions & 10 deletions profiles/default/standards/testing/unit-tests.md

This file was deleted.

6 changes: 3 additions & 3 deletions profiles/default/workflows/implementation/implement-task.md
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,6 @@ Guide your implementation using:
- **The existing patterns** that you've found and analyzed.
- **User Standards & Preferences** which are defined below.

Self-verify and test your work:
- IF your tasks direct you to write tests, ensure all of the tests you've written pass.
- Double-check, test, or view the elements you've implemented to verify they are all present and in working order before reporting on your implementation.
Self-verify and test your work by:
- Running ONLY the tests you've written (if any) and ensuring those tests pass.
- IF your task involves user-facing UI, and IF you have access to browser testing tools, open a browser and use the feature you've implemented as if you are a user to ensure a user can use the feature in the intended way.
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
1. **Analyze YOUR assigned task:** Take note of the specific task and sub-tasks that have been assigned to your role. Do NOT implement task(s) that are assigned to other roles.
2. **Search for existing patterns:** Find and state patterns in the codebase and user standards to follow in your implementation.
3. **Implement accoding to requirements & standards:** Implement your tasks by following your provided tasks, spec and ensuring alignment with "User's Standards & Preferences Compliance" and self-test and verify your own work.
3. **Implement according to requirements & standards:** Implement your tasks by following your provided tasks, spec and ensuring alignment with "User's Standards & Preferences Compliance".
4. **Update tasks.md with your tasks status:** Mark the task and sub-tasks in `tasks.md` that you've implemented as complete by updating their checkboxes to `- [x]`
5. **Document your implementation:** Create your implementation report in this spec's `implementation` folder detailing the work you've implemented.
Original file line number Diff line number Diff line change
@@ -1,8 +1,8 @@
1. **Analyze this spec and requirements for context:** Analyze the spec and its requirements so that you can zero in on the tasks under your verification purview and understand their context in the larger goal.
2. **Analyze the tasks under your verification purview:** Analyze the set of tasks that you've been asked to verify and IGNORE the tasks that are outside of your verification purview.
3. **Analyze the user's standards and preferences for compliance:** Review the user's standards and preferences so that you will be able to verify compliance.
4. **Run the tests that were written for the tasks under your verification purview:** Verify how many are passing and failing.
5. **(if applicable) view the implementation in a browser:** If your verification purview involves UI implementations, open a browser to view, verify and take screenshots.
4. **Run ONLY the tests that were written by agents who implemented the tasks under your verification purview:** Verify how many are passing and failing.
5. **(if applicable) view the implementation in a browser:** If your verification purview involves UI implementations, open a browser to view, verify and take screenshots and store screenshot(s) in `agent-os/specs/[this-spec]/verification/screenshots`.
6. **Verify tasks.md status has been updated:** Verify and ensure that the tasks in `tasks.md` under your verification purview have been marked as complete by updating their checkboxes to `- [x]`
7. **Verify that implementations have been documented:** Verify that the implementer agent(s) have documented their work in this spec's `implementation` folder.
8. **Document your verification report:** Write your verification report in this spec's `verification` folder.
7. **Verify that implementations have been documented:** Verify that the implementer agent(s) have documented their work in this spec's `agent-os/specs/[this-spec]/implementation`. folder.
8. **Document your verification report:** Write your verification report in this spec's `agent-os/specs/[this-spec]/verification`. folder.
Loading