commit 95ea24e4c91a9073da80a4739828cf06819a752d
parent 99f949f8afcffdc1b4bd06eebf223e1dc275e4a4
Author: andrewlaack-collab <andrew.laack@imbue.com>
Date: Thu, 5 Feb 2026 20:15:02 +0000
Update 'agentic' usage guidelines (#16)
* Vibecoding UX considerations :(
* Bug fix for session loading. OpenCode works!
* Bug fix for session loading. OpenCode works!
* Stronger wording
* Updated wording
* Refactoring
* Black
* Added install stuff
* Reworded skill
* Updeated some more usage stuff
* Rewording
Diffstat:
2 files changed, 29 insertions(+), 2 deletions(-)
diff --git a/README.md b/README.md
@@ -24,6 +24,33 @@ Compare against a base ref/commit:
vet "Refactor storage layer" --base-commit main
```
+## Using Vet with Coding Agents
+
+Vet ships as an [agent skill](https://agentskills.io) that coding agents like [OpenCode](https://opencode.ai) and [Codex](https://github.com/openai/codex) can discover and use automatically. When installed, agents will proactively run vet after code changes and include conversation history for better analysis.
+
+### Install the skill globally
+
+```bash
+mkdir -p ~/.agents/skills/vet/scripts && \
+ curl -fsSL https://raw.githubusercontent.com/imbue-ai/vet/main/skills/vet/SKILL.md \
+ -o ~/.agents/skills/vet/SKILL.md && \
+ curl -fsSL https://raw.githubusercontent.com/imbue-ai/vet/main/skills/vet/scripts/export_opencode_session.py \
+ -o ~/.agents/skills/vet/scripts/export_opencode_session.py && \
+ curl -fsSL https://raw.githubusercontent.com/imbue-ai/vet/main/skills/vet/scripts/export_codex_session.py \
+ -o ~/.agents/skills/vet/scripts/export_codex_session.py
+```
+
+This places the skill in `~/.agents/skills/vet/`, which is discovered by both OpenCode and Codex.
+
+### Install per-project
+
+To have agents use vet automatically in a specific repo, copy the skill into the project:
+
+```bash
+cp -r /path/to/vet/skills/vet .agents/skills/vet
+git add .agents/skills/vet && git commit -m "Add vet agent skill"
+```
+
## How it works
Vet snapshots the repo and diff, optionally adds a goal and agent conversation, runs LLM checks, then filters/deduplicates findings into a final list of issues.
diff --git a/skills/vet/SKILL.md b/skills/vet/SKILL.md
@@ -34,13 +34,13 @@ Always include conversation history for best results. Set the session environmen
**OpenCode:**
```bash
export VET_SESSION_ID="<current-session-id>"
-vet "goal" --history-loader "python /path/to/skills/vet/scripts/export_opencode_session.py"
+vet "goal" --history-loader "python ~/.agents/skills/vet/scripts/export_opencode_session.py"
```
**Codex:**
```bash
export CODEX_SESSION_FILE="<path-to-current-session.jsonl>"
-vet "goal" --history-loader "python /path/to/skills/vet/scripts/export_codex_session.py"
+vet "goal" --history-loader "python ~/.agents/skills/vet/scripts/export_codex_session.py"
```
### Finding Your Session