# sfagent-tools

This plugin runs test conversations against your Agentforce agent, checks routing and guardrails, and returns a scored report.

- plugin · `type=plugin` · Claude Code · ★ 24
- Author: lucianostraga
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/lucianostraga/sfagent-tools
- Updated: 2026-05-26
- HTML: https://affpapa.org/en/nevibecode/plugins/sfagent-tools

## What it does

The plugin pulls your agent's configuration from the org — subagents, actions, descriptions — so it knows what to exercise. Then it drives multi-turn sessions headlessly through `sf agent preview`, and you can follow the markdown transcript live in a split pane. Injection attempts, off-topic asks and manager-escalation triggers get probed along the way.

Put your expectations in `sfagent-config.yaml`: per-subagent rules, global rules, custom scenarios. Each one comes back as pass or fail, and the run ends with a report scoring routing accuracy, guardrail strength and multi-turn coherence. A YAML export for `sf agent test run-eval` lets the same scenarios run on every commit.

No new credentials are involved. The plugin reuses your `sf` CLI auth, so `sf org login web` once is enough. Production orgs are blocked at the tool level, which means tests only touch sandboxes, scratch orgs or Developer Edition. Diagnostics rely on `sf agent trace`, so routing and action details need `sf` CLI 2026-05-20 or later.

## Requirements
- Works with: Salesforce

## Install

### Claude Code (native platform)

**Plugin marketplace**
In the agent chat (not a shell):
```text
/plugin marketplace add anthropics/claude-plugins-community
```
In the agent chat (not a shell):
```text
/plugin install sfagent-tools@claude-community
```

## When to use it
- check whether a customer request reaches the right subagent
- catch attempts to talk the agent out of its role
- compare replies against rules in `sfagent-config.yaml`
- see if context survives a multi-turn conversation
- export scenarios as YAML for CI runs

## Example prompts
- Test my Agentforce agent
- List the subagents and actions in this org
- Run the scenarios from `sfagent-config.yaml` and show which rules fail

## Similar
- [example-skills](https://affpapa.org/en/nevibecode/plugins/example-skills.md) — plugin · ★ 183839
- [superpowers](https://affpapa.org/en/nevibecode/plugins/superpowers.md) — plugin · ★ 305248
- [ponytail](https://affpapa.org/en/nevibecode/plugins/ponytail.md) — plugin · ★ 176443
- [mattpocock-skills](https://affpapa.org/en/nevibecode/plugins/mattpocock-skills.md) — plugin · ★ 303516
- [promptfoo-evals](https://affpapa.org/en/nevibecode/plugins/promptfoo-evals.md) — plugin · ★ 26372
- [skill-creator](https://affpapa.org/en/nevibecode/plugins/skill-creator.md) — plugin · ★ 44635

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md