Compress prompts, reuse local cache, and forecast token cost before they reach Copilot.
📝0memory🔗0graph💾0cache🔌0peers📚0studied
Optimized
Packing local context and consulting the semantic cache...
Copilot response
streaming
Cost
$0.00000
Saved
—
Tokens
—
Est. out
—
Cache
—
Pricing
—
▶Refinements0
▶Credit Forecast (Monthly)Included
⚠️ Values shown are estimates based on your configured Copilot plan, model premium-request multiplier, and your estimated request volume — they are not actual billed amounts. Adjust the plan, requests/day, and overage price in settings.
Item
Value
Projected monthly overage
$0.00
▶Lint Diagnostics0
🔒 Secret Detection
Built-in patterns (read-only)
Custom patterns
{{SECRET_PATTERN_HELP_TOOLTIP}}
✓ Settings saved
🤖 Agents
Add a custom agent skill for this workspace, or manage the bundled SDLC agents. New agents hot-reload on the next prompt and run via /<id>.
Create a new agent
📄Drag & drop a .md / .txt file here, or
✓ Agent created
📋 Instructions Manager
Every directive the AI currently sees, grouped by source. Click a source to open the file.
When rules disagree, higher-priority sources win. Order, highest authority first:
Contradictions detected between active directives, with a suggested resolution.
Git history (author + date) for a chosen instruction file.
SDLC personas bundled with this build. Enable one to install it as a workspace skill the AI can adopt.
Paste a task to preview which directives apply and whether any active conflicts would affect it.
{{EXTRA_JS_SCRIPTS}}
🔒Local until you send. Prompts stay on your machine; only the Copilot call you trigger leaves the editor.