Skip to content

Commit d628682

Browse files
TyDunnTyDunn
authored
Merge pull request #8509 from continuedev/bekah/agent-docs
docs: Add comprehensive agent documentation
2 parents b1cd3e9 + 4ca6ca3 commit d628682

File tree

8 files changed

+599
-41
lines changed

8 files changed

+599
-41
lines changed

‎docs/docs.json‎

Lines changed: 6 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -56,8 +56,12 @@
5656
},
5757
{
5858
"group": "Agents",
59-
"icon": "cloud",
60-
"pages": ["hub/agents/intro"]
59+
"icon": "robot",
60+
"pages": [
61+
"hub/agents/intro",
62+
"hub/agents/overview",
63+
"hub/agents/create-and-edit"
64+
]
6165
},
6266
"hub/sharing",
6367
"hub/source-control"

‎docs/hub/agents/create-and-edit.mdx‎

Lines changed: 261 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,261 @@
1+
---
2+
title: "Create and Edit Agents"
3+
description: "Build custom AI workflows with prompts, rules, and tools through the Continue Hub interface"
4+
sidebarTitle: "Create & Edit"
5+
---
6+
7+
<Card title="Creating Custom Agents" icon="wrench">
8+
You can only create and edit Agents through Mission Control in the Continue Hub web interface.
9+
This ensures proper validation, versioning, and team collaboration features.
10+
</Card>
11+
12+
## Try an Agent First: Your 60-Second Challenge
13+
14+
Before creating your own agent, let's see one in action! The fastest way to experience the power of Continue agents is with our demo repository.
15+
16+
![Agent Mission Control](/images/hub/assets/images/agent-mission-control.gif)
17+
18+
### See an Agent Create a Pull Request in Under 60 Seconds
19+
20+
<Steps>
21+
<Step title="🍴 Fork Our Demo Repository">
22+
Get a safe sandbox to experiment with:
23+
24+
**Option 1: GitHub Web Interface**
25+
Visit [github.com/continuedev/demo-project](https://github.com/continuedev/demo-project) and click **Fork**
26+
27+
**Option 2: GitHub CLI**
28+
```bash
29+
gh repo fork continuedev/demo-project
30+
```
31+
</Step>
32+
33+
<Step title="🎯 Access Mission Control">
34+
Go to [hub.continue.dev/agents](https://hub.continue.dev/agents) and:
35+
- **Connect GitHub** and authorize Continue when prompted
36+
- This gives agents access to create PRs in your repositories
37+
</Step>
38+
39+
<Step title="🚀 Run Your First Agent">
40+
**Target your forked demo repo** and try one of these commands:
41+
42+
<Tabs>
43+
<Tab title="Add New Feature">
44+
```
45+
Create a new function that calculates fibonacci numbers.
46+
```
47+
</Tab>
48+
49+
<Tab title="Fix Existing Bug">
50+
```
51+
Fix the TypeError in api/users.ts
52+
```
53+
</Tab>
54+
55+
<Tab title="Add Documentation">
56+
```
57+
Add comprehensive README documentation with setup instructions.
58+
```
59+
</Tab>
60+
</Tabs>
61+
</Step>
62+
63+
<Step title="✨ Watch the Magic">
64+
**Result:** A fully-formed Pull Request opens in your demo repo with:
65+
- Complete implementation or fix
66+
- Proper commit messages
67+
- Detailed PR description
68+
- Ready for review and merge
69+
</Step>
70+
</Steps>
71+
72+
<Info>
73+
**Why Use the Demo Repo?**
74+
- **Safe testing environment** - No risk to your production code
75+
- **Pre-configured issues** - Common bugs and features to practice with
76+
- **Immediate results** - See agents in action without setup complexity
77+
- **Learn by example** - Study the generated code and PR descriptions
78+
</Info>
79+
80+
<Tip>
81+
Once you see how agents work with the demo repo, you'll understand exactly how to create and customize your own for real projects!
82+
</Tip>
83+
84+
---
85+
86+
## Creating an Agent
87+
88+
![Create an agent gif](/images/hub/create-an-agent.gif)
89+
90+
91+
<Steps>
92+
<Step title="🧭 Navigate to Create Agent">
93+
From the Continue Hub top navigation bar, select **"+"** → **"New Agent"**.
94+
95+
This opens the **Create an Agent (beta)** form with all required fields.
96+
97+
</Step>
98+
99+
<Step title="🧩 Fill Out the Agent Form">
100+
Configure your agent with the following fields:
101+
102+
| Field | What to Enter | Example |
103+
|-------|---------------|---------|
104+
| **Name** | Display name shown in the Hub | `GitHub PR Agent` |
105+
| **Prompt** | First instruction the agent receives | `Open a GitHub PR to fix the specified issue.` |
106+
| **Description** | What the agent does | `Creates a pull request and includes AI-generated summaries.` |
107+
| **Tools (MCPs)** | Select built-in or custom MCPs | `GitHub, PostHog, Supabase` |
108+
| **Rules** | Add any organizational rules | `continuedev/gh-pr-commit-workflow, continuedev/summarization` |
109+
| **Model** | Choose a default LLM | `Claude Sonnet 4.5` |
110+
| **Owner + Slug** | Determines namespace | `my-org/github-pr-agent` |
111+
| **Visibility** | Access level | `Public, Organization, or Private` |
112+
113+
<Info>
114+
**Tip**: Start with a simple prompt and add complexity through rules and tools. The prompt should be a clear, single instruction that defines the agent's primary goal.
115+
</Info>
116+
</Step>
117+
118+
<Step title="💡 Preview the Configuration">
119+
As you fill in the fields, the right-hand panel shows a live preview of your agent's YAML configuration:
120+
121+
```yaml
122+
name: Open PR with Fix
123+
description: Open a GitHub PR to fix the specified issue
124+
tools: built_in, anthropic/github-mcp
125+
rules:
126+
- continuedev/gh-pr-commit-workflow
127+
- continuedev/summarization
128+
model: Claude Sonnet 4.5
129+
visibility: public
130+
```
131+
132+
This preview helps you verify the configuration before creating the agent.
133+
</Step>
134+
135+
<Step title="✅ Create and Test">
136+
Click **"Create Agent"** to save and publish it.
137+
138+
Your agent is immediately available to run in:
139+
- Mission Control web interface
140+
- TUI mode: `cn --agent your-org/your-agent-name`
141+
- Headless mode: `cn --agent -p your-org/your-agent-name "prompt" --auto`
142+
</Step>
143+
</Steps>
144+
145+
## Editing an Agent
146+
147+
You can edit any agent you own or that belongs to your organization.
148+
149+
<Steps>
150+
<Step title="🧭 Access the Agent">
151+
From the **Agents** page or Mission Control view, click your agent's name, then select **"Edit Agent"**.
152+
</Step>
153+
154+
<Step title="🔧 Update Configuration">
155+
You can modify any of these components:
156+
157+
<Tabs>
158+
<Tab title="Prompt">
159+
**Refine task behavior**
160+
161+
```
162+
# Before
163+
Fix the bug in the authentication system
164+
165+
# After
166+
Fix the authentication bug by:
167+
1. Identifying the root cause
168+
2. Implementing a secure solution
169+
3. Adding appropriate tests
170+
4. Opening a PR with detailed explanation
171+
```
172+
</Tab>
173+
174+
<Tab title="Rules">
175+
**Enforce team standards**
176+
177+
Add or remove organizational rules:
178+
- `continuedev/security-first` - Always scan for vulnerabilities
179+
- `continuedev/test-coverage` - Require tests for new code
180+
- `my-org/style-guide` - Follow company coding standards
181+
</Tab>
182+
183+
<Tab title="Tools (MCPs)">
184+
**Connect additional systems**
185+
186+
- Add GitHub MCP for repository operations
187+
- Include Sentry MCP for error monitoring
188+
- Connect Supabase MCP for database operations
189+
- Use custom MCPs for internal tools
190+
</Tab>
191+
192+
<Tab title="Model & Visibility">
193+
**Adjust behavior and access**
194+
195+
- **Model**: Switch between Claude, GPT-4, or other supported LLMs
196+
- **Visibility**: Control who can see and use the agent
197+
- Public: Anyone can discover and use
198+
- Organization: Only your team members
199+
- Private: Only you can access
200+
</Tab>
201+
</Tabs>
202+
</Step>
203+
204+
<Step title="💾 Save Changes">
205+
Click **"Update Agent"** when finished.
206+
207+
The updated version is instantly available to your team with automatic versioning for change tracking.
208+
</Step>
209+
</Steps>
210+
211+
<Warning>
212+
Test your agent thoroughly to ensure they interact as expected.
213+
</Warning>
214+
215+
## Example Agent Configurations
216+
217+
Here are proven agent configurations you can create or use as inspiration:
218+
219+
<CardGroup cols={2}>
220+
<Card title="Security Scanner Agent" icon="shield" href="https://hub.continue.dev/continuedev/snyk-continuous-ai-agent">
221+
**Snyk Continuous AI Agent** - Comprehensive security scanning with Snyk MCP integration.
222+
223+
Automates dependency analysis, vulnerability scanning, and creates remediation PRs with AI-powered fix suggestions.
224+
</Card>
225+
226+
<Card title="Performance Monitor" icon="gauge" href="https://hub.continue.dev/continuedev/netlify-continuous-ai-agent">
227+
**Netlify Continuous AI Agent** - Performance optimization with A/B testing and monitoring.
228+
229+
Tracks Core Web Vitals, identifies regressions, and provides optimization recommendations.
230+
</Card>
231+
232+
<Card title="Error Monitoring Agent" icon="bug" href="https://hub.continue.dev/continuedev/sentry-continuous-ai-agent">
233+
**Sentry Continuous AI Agent** - Automated error analysis and issue creation.
234+
235+
Monitors production errors, provides root cause analysis, and creates actionable GitHub issues.
236+
</Card>
237+
238+
<Card title="GitHub Management" icon="github" href="https://hub.continue.dev/continuedev/github-manager-ai-agent">
239+
**GitHub Manager AI Agent** - Comprehensive GitHub workflow automation.
240+
241+
Handles issue triage, PR reviews, and release note generation with natural language prompts.
242+
</Card>
243+
244+
<Card title="Database Management" icon="database" href=" https://hub.continue.dev/continuedev/supabase-agent">
245+
**Supabase Continuous AI Agent** - Database security and management workflows.
246+
247+
Audits Row Level Security, identifies vulnerabilities, and generates fixes automatically.
248+
</Card>
249+
</CardGroup>
250+
251+
252+
## Troubleshooting
253+
254+
Common issues and solutions:
255+
256+
| Problem | Solution |
257+
|---------|----------|
258+
| **Agent doesn't complete tasks** | Simplify the prompt, add more specific instructions, verify tool permissions |
259+
| **Tools aren't working** | Check MCP configuration, verify API keys and permissions in Hub settings |
260+
| **Inconsistent behavior** | Add rules to enforce consistent patterns, test with various input scenarios |
261+
| **Performance issues** | Consider model choice, simplify complex multi-step workflows, optimize tool usage |

0 commit comments

Comments
 (0)