Skip to content

feat: add help subcommand to /fullsend skill #6

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
durandom merged 1 commit into from
Jun 11, 2026
Merged

feat: add help subcommand to /fullsend skill #6

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions .claude/skills/fullsend/SKILL.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,9 @@ description: |
watch or monitor a fullsend run, comment on an issue, or manage labels.
Also use when asked to run sandbox diagnostics, debug the sandbox environment,
check if yarn/corepack/openspec work, or build a custom fullsend agent.
Also use when asked what fullsend agents are, how they work, how to use them,
what agents are available, how to get started with fullsend, or how the local
deployment is configured.
---

# /fullsend
Expand Down Expand Up @@ -91,6 +94,7 @@ To add a variable, create an env file and wire it via `host_files` in the harnes
| `debug <#issue> [--repo]` | Run sandbox diagnostics (shortcut for `trigger debug`) |
| `comment <#issue> <message> [--repo]` | Post a comment on an issue or PR |
| `label <#issue> <add\|remove> <label> [--repo]` | Add or remove a label on an issue or PR |
| `help [topic]` | Onboarding companion — agent pipeline, local deployment overview, upstream docs |

If no arguments are given, display this table and ask which the user wants.

Expand All @@ -111,6 +115,7 @@ Parse the first word after `/fullsend` as the subcommand.
| `debug` | `references/debug.md` |
| `comment` | `references/comment.md` |
| `label` | `references/label.md` |
| `help` | `references/help.md` |

</routing>

Expand Down
198 changes: 198 additions & 0 deletions .claude/skills/fullsend/references/help.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,198 @@
# help

Context-aware onboarding companion for fullsend. Shows the agent pipeline, inspects the local deployment's customizations, and links to upstream docs.

## Usage

```
/fullsend help [topic]
```

- No argument: show all sections
- `agents`: agent pipeline and catalog
- `setup` or `local`: getting started / running locally
- `customization` or `config`: how to customize agents, skills, harness

## Procedure

### 1. Resolve the target repo

Follow the repo resolution order from the SKILL.md `<repo_resolution>` section to find the repo with `.fullsend/` configuration.

If in a monorepo (e.g., `rhdh-plugins`), note this — the agent operates at repo level, not workspace level.
Comment on lines +18 to +22

### 2. Route by topic

| Argument | Sections to show |
|----------|-----------------|
| _(none)_ | All: pipeline → local deployment → upstream links |
| `agents` | Pipeline + agent catalog only |
| `setup` / `local` | Getting started links + prerequisites only |
| `customization` / `config` | Local deployment inspection + customization links |

### 3. Agent pipeline (topic: `agents` or all)

Display this pipeline diagram:

```
┌─────────────────── Fullsend Agent Pipeline ───────────────────┐
│ │
│ issue opened │
│ │ │
│ ▼ │
│ triage ─── /fs-triage │
│ │ │
│ │ adds `ready-to-code` label (auto for bugs/docs/perf) │
│ ▼ │
│ code ──── /fs-code │
│ │ │
│ │ pushes branch, creates PR, adds `ready-for-review` │
│ ▼ │
│ review ─── /fs-review │
│ │ │
│ ├── approved ──────────────────────▶ merge │
│ │ │ │
│ └── changes requested ▼ │
│ │ retro ── /fs-retro │
│ ▼ │
│ fix ──── /fs-fix │
│ │ │
│ └── back to review (max iterations capped) │
│ │
│ Scoring (independent): prioritize ── /fs-prioritize │
│ Diagnostics: debug ──────── /fs-debug │
│ │
└───────────────────────────────────────────────────────────────┘
```

Then show the agent catalog:

| Agent | Trigger | What it does |
|-------|---------|-------------|
| **triage** | New issue opened, `/fs-triage` | Assesses issue, asks clarifying questions, labels `ready-to-code` for bugs/docs/perf |
| **code** | `ready-to-code` label, `/fs-code` | Implements fix/feature in sandbox, commits to feature branch |
| **review** | `ready-for-review` label, `/fs-review` | Multi-dimension code review (correctness, security, style, docs) |
| **fix** | Review requests changes, `/fs-fix` | Addresses specific review findings, pushes fix commit |
| **retro** | PR merged/closed, `/fs-retro` | Retrospective on the full workflow — files improvement issues |
Comment on lines +74 to +76
| **prioritize** | Schedule or `/fs-prioritize` | RICE scoring for project board ranking |
| **debug** | `/fs-debug` | Sandbox diagnostics — checks toolchain, network, env |

All agents run in a sandboxed container (OpenShell) with no direct push access. Pre/post scripts on the runner handle git operations and safety gates (secret scanning, protected path checks).

### 4. Local deployment inspection (topic: `customization` or all)

Read the target repo's `.fullsend/` directory and present a deployment summary. This section is dynamic — it reflects the actual state of the repo's configuration.

#### 4a. Read config.yaml

```bash
cat <repo>/.fullsend/config.yaml
```

Show which roles are enabled:
```
Enabled roles: triage, coder, review, fix
```

#### 4b. Scan customized directory

```bash
ls <repo>/.fullsend/customized/harness/*.yaml 2>/dev/null
ls <repo>/.fullsend/customized/agents/*.md 2>/dev/null
ls <repo>/.fullsend/customized/skills/*/SKILL.md 2>/dev/null
ls <repo>/.fullsend/customized/env/*.env 2>/dev/null
ls <repo>/.fullsend/customized/policies/*.yaml 2>/dev/null
Comment on lines +100 to +104
```

#### 4c. For each customized harness, read it and summarize the overrides

For each `*.yaml` in `customized/harness/`, read the file and extract:
- `image` — what custom image is used (and why it differs from upstream)
- `skills` — what skills are loaded (especially custom ones)
- `host_files` — what extra env/config files are mounted
- `policy` — which network policy is applied
- `timeout_minutes` — if different from upstream default

Present as a per-agent summary:

```
## Your Deployment

### code agent (customized)
Image: ghcr.io/redhat-developer/rhdh-fullsend-code:latest
Skills: code-implementation, monorepo-workspace-routing (custom)
Env: gcp-vertex.env, code-agent.env, rhdh-toolchain.env (custom), yarn-proxy.env (custom)
Policy: policies/code.yaml (custom — adds repo.yarnpkg.com, corepack binary)
Timeout: 35 min
Prompt: agents/code.md (customized — adds monorepo routing, zero-trust principle)

### fix agent (customized)
Image: ghcr.io/redhat-developer/rhdh-fullsend-code:latest
Skills: fix-review, monorepo-workspace-routing (custom)
...

### triage agent (upstream defaults)
### review agent (upstream defaults)
```

Mark items as "(custom)" when they exist in `customized/` and differ from the upstream scaffold default. For agents without customizations, show as "upstream defaults."

#### 4d. Custom skills

For each skill in `customized/skills/`, read the `SKILL.md` and show name + description:

```
### Custom Skills
- monorepo-workspace-routing — Navigate to the correct workspace in a monorepo before starting work
```

#### 4e. Custom env files

For each `.env` in `customized/env/`, show what it sets:

```
### Custom Env Files
- rhdh-toolchain.env — sets COREPACK_HOME=/tmp/corepack (writable dir for corepack)
- yarn-proxy.env — maps HTTP_PROXY → YARN_HTTP_PROXY (yarn ignores standard proxy vars)
```
Comment on lines +149 to +157

### 5. Upstream docs (topic: `setup` or all)

Present as a linked reference table. Construct URLs as `https://github.com/fullsend-ai/fullsend/blob/main/<path>`.

Comment on lines +159 to +162
```
## Learn More

### Getting Started
- [Installation & setup](https://github.com/fullsend-ai/fullsend/blob/main/docs/guides/getting-started/installation.md)
- [GitHub setup](https://github.com/fullsend-ai/fullsend/blob/main/docs/guides/getting-started/github-setup.md)
- [Running agents locally](https://github.com/fullsend-ai/fullsend/blob/main/docs/guides/user/running-agents-locally.md)

### Agent Reference
- [Agent catalog](https://github.com/fullsend-ai/fullsend/blob/main/docs/agents/README.md)
- [Code agent](https://github.com/fullsend-ai/fullsend/blob/main/docs/agents/code.md)
- [Triage agent](https://github.com/fullsend-ai/fullsend/blob/main/docs/agents/triage.md)
- [Review agent](https://github.com/fullsend-ai/fullsend/blob/main/docs/agents/review.md)
- [Fix agent](https://github.com/fullsend-ai/fullsend/blob/main/docs/agents/fix.md)
- [Prioritize agent](https://github.com/fullsend-ai/fullsend/blob/main/docs/agents/prioritize.md)
- [Retro agent](https://github.com/fullsend-ai/fullsend/blob/main/docs/agents/retro.md)

### Customization
- [Customizing agents (harness config)](https://github.com/fullsend-ai/fullsend/blob/main/docs/guides/user/customizing-agents.md)
- [Customizing with AGENTS.md](https://github.com/fullsend-ai/fullsend/blob/main/docs/guides/user/customizing-with-agents-md.md)
- [Customizing with skills](https://github.com/fullsend-ai/fullsend/blob/main/docs/guides/user/customizing-with-skills.md)
- [Bugfix workflow](https://github.com/fullsend-ai/fullsend/blob/main/docs/guides/user/bugfix-workflow.md)

### Architecture
- [Architecture overview](https://github.com/fullsend-ai/fullsend/blob/main/docs/architecture.md)
- [Glossary](https://github.com/fullsend-ai/fullsend/blob/main/docs/glossary.md)
- [Vision](https://github.com/fullsend-ai/fullsend/blob/main/docs/vision.md)
```

### 6. Suggest next steps

Based on the context, suggest relevant `/fullsend` commands:

- "Run `/fullsend validate` to check your customizations against upstream for drift."
- "Run `/fullsend trigger code #<N>` to trigger the code agent on an issue."
- "Run `/fullsend inspect <run-id>` to investigate a specific agent run."