@mrclrchtr/supi-agent
Profile-based foreground Agent Runs for the Pi coding agent
Package details
Install @mrclrchtr/supi-agent from npm and Pi will load the resources declared by the package manifest.
$ pi install npm:@mrclrchtr/supi-agent- Package
@mrclrchtr/supi-agent- Version
7.3.0- Published
- Sep 24, 2026
- Downloads
- 1,512/mo · 713/wk
- Author
- mrclrchtr
- License
- MIT
- Types
- extension
- Size
- 52.8 MB
- Dependencies
- 3 dependencies · 4 peers
Pi manifest JSON
{
"image": "https://raw.githubusercontent.com/mrclrchtr/supi/main/packages/supi-agent/assets/social-preview.png",
"extensions": [
"./src/extension.ts"
]
}Security note
Pi packages can execute code and influence agent behavior. Review the source before installing third-party packages.
README
@mrclrchtr/supi-agent
Profile catalogue, field-level profile settings, and resource policy for foreground Agent Runs in PI.
This package is the policy layer for @mrclrchtr/supi-agent-runtime. It discovers Profile Directory sources from package defaults, ~/.pi/agent/supi/agents/, and trusted project .pi/supi/agents/ directories. It also contributes per-profile Model and Thinking rows to /supi-settings.
Built-in profiles:
explore— read and headless Code Intelligence tools, no instruction files, read-only.general— read, bash, edit, write, and headless Code Intelligence tools, global/project instruction files, mutation-capable.
Profile sources overlay fields by ID with project → global → package precedence. A partial user manifest can pin only model or thinking; package tools and prompts continue to flow through. An invalid source falls through with a bounded diagnostic. Profiles are context-isolated, not permission- or filesystem-sandboxed.
Changes to Agent Run tool and each profile's Model and Thinking settings apply immediately. /agents stays available when agent_run is off.
Agent Run row
The agent_run transcript row shows the effective model of a Delegation Task as model: provider/model-id. It adds thinking: <level> when the run reports a thinking level. The live progress row, the collapsed result, and the expanded result show these facts first, followed by turns, tool uses, and Usage. A run without model data shows no model segment.
/agents
Use /agents in TUI mode to inspect active Agent Runs, every completed Delegation Batch in the current parent session, effective Agent Profiles, and bounded Profile Diagnostics. The Runs view keeps the bounded Conversation View separate from a full human-only transcript. It shows status, model, thinking level, turns, tool uses, Usage, safe tool activity, steering, assistant text, and retention notices. Transcript messages stay outside parent tool results.
- Use Tab, Shift+Tab, or Left/Right to change sections.
- Use Up/Down to select a run, profile, or diagnostic.
- At terminal widths of 100 columns or more, the Runs view shows a run list and transcript side by side. On narrower terminals, use Left to open the run list, Up/Down to select a run, and Enter or Right to open its transcript.
- The viewer uses the full terminal. The run list and controls stay visible while the transcript scrolls. Mouse-wheel input scrolls the transcript or selects a run.
- The viewer follows new transcript output by default (
LIVE). Use Page Up/Page Down to scroll through wrapped lines. Page Up pauses auto-scroll (PAUSED); live updates keep the reading position. The status shows how many lines are below the view, not an unread count. - Press Home to read task metadata and earlier output. Press End to resume auto-scroll. Page Down also resumes auto-scroll when it reaches the end.
- Press
fto pause or resume auto-scroll. Selecting a different run resumes auto-scroll. - Press
tor Ctrl+T to hide or show thinking. Pressoor Ctrl+O to show or hide tool input and output. The Ctrl shortcuts match PI defaults. You can also click a raw tool block to show or hide its input and output. These shortcuts do not apply while you enter a steering message. - The bounded Conversation View can omit old entries. The human-only transcript keeps all captured messages for every batch in the current parent session.
- Press
sto steer the selected running Agent Run. A text field opens inside the overlay; press Enter to send or Esc to cancel. - Press
xto request a stop for the selected starting or running Agent Run, then press Enter oryto confirm. A startup stop can wait for PI setup to finish. - Press Esc to close the viewer. This does not stop Agent Runs. PI's normal outer-tool cancellation still stops the full Delegation Batch.
Transcript files are temporary JSONL files in a private directory for the parent session. They have no application cap or restart recovery. A storage failure marks capture incomplete but does not stop a run. The viewer shows the available transcript and the incomplete status. Session shutdown removes the directory.
Other PI modes show only an unavailable notice. The overlay does not change global tuiMode or add a text control protocol, persistent child history, or replay files. Closing the viewer does not stop runs.
