agent-catalog-new 0.2.0__tar.gz

agent_catalog_new-0.2.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Ervin
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

agent_catalog_new-0.2.0/PKG-INFO

@@ -0,0 +1,201 @@
+Metadata-Version: 2.4
+Name: agent-catalog-new
+Version: 0.2.0
+Summary: Declarative agent registry — catalog, discover, and diff agent capabilities across environments
+Author-email: Ervin <ervin@nexusgate.tech>
+License: MIT
+Keywords: agents,catalog,registry,ai-agents,agent-management,mcp
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Software Development :: Libraries
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Provides-Extra: dev
+Requires-Dist: pytest<10,>=9.0; extra == "dev"
+Requires-Dist: pytest-cov<8,>=7.0; extra == "dev"
+Requires-Dist: ruff<1,>=0.15; extra == "dev"
+Requires-Dist: coverage[toml]<8,>=7.6; extra == "dev"
+Requires-Dist: mypy>=1.15; extra == "dev"
+Provides-Extra: mcp
+Requires-Dist: mcp<2,>=1.0; extra == "mcp"
+Dynamic: license-file
+
+# Agent Catalog
+
+[![CI](https://github.com/ewardGPT/agent-catalog/actions/workflows/ci.yml/badge.svg)](https://github.com/ewardGPT/agent-catalog/actions/workflows/ci.yml)
+[![PyPI](https://img.shields.io/pypi/v/agent-catalog)](https://pypi.org/project/agent-catalog/)
+[![Python](https://img.shields.io/pypi/pyversions/agent-catalog)](https://pypi.org/project/agent-catalog/)
+[![License](https://img.shields.io/github/license/ewardGPT/agent-catalog)](LICENSE)
+
+A registry for AI agents. Track every agent you own — what it does, what tools it uses, what model it runs, where it's deployed. YAML manifests stored on disk. No database. Git-ops friendly.
+
+Think `npm` for agents. One tool to list, search, diff, and invoke them all.
+
+```bash
+pip install agent-catalog
+agent-catalog register ./my-agent.yaml
+agent-catalog list
+```
+
+## Why
+
+You have agents scattered across repos, environments, and projects. Prod vs staging drift goes unnoticed. A dependency changes and nothing tells you. You grep through directories to find which agent has the capability you need.
+
+Agent Catalog fixes that. One directory is the source of truth. Every command queries it.
+
+## Features
+
+- **Registry.** Register agents from YAML manifests. List, search, get details. Filter by environment, capability, tool, surface.
+- **Diff.** Compare manifests across environments (`--left-env staging --right-env production`). Catch drift before it hits prod.
+- **Consistency.** `agent-catalog doctor` finds orphaned files and missing manifests.
+- **Security audit.** Flags MCP endpoints without auth, write capabilities without confirmation gates, idempotent tools that should not be.
+- **MCP server.** Expose the catalog as MCP tools — any MCP client can discover agents and invoke their capabilities.
+- **HTTP dashboard.** `agent-catalog serve` starts a web UI on `:8420`.
+- **Python SDK.** `from agent_catalog.client import CatalogClient` — sync or async.
+- **Decorator API.** Define agents inline with `@agent`, `@capability`, `@tool`. Manifests auto-generated from Python classes.
+- **Runtime loader.** Load an agent class from the catalog and invoke a capability from the command line or API.
+
+## Quick start
+
+```bash
+pip install agent-catalog
+
+# Register an agent
+agent-catalog register ./examples/agentic-inbox.yaml
+
+# See everything you have
+agent-catalog list
+
+# Get details
+agent-catalog get agentic-inbox
+
+# Search
+agent-catalog search --capability send_email
+
+# Validate a manifest without registering
+agent-catalog validate ./my-agent.yaml
+```
+
+## Walkthrough: managing agents across environments
+
+```bash
+# Register your staging agent
+agent-catalog register ./staging/agent.yaml --env staging
+
+# Later, register production
+agent-catalog register ./prod/agent.yaml
+
+# See the drift
+agent-catalog diff agentic-inbox --left-env staging --right-env production
+
+# Validate your changes
+agent-catalog validate ./prod/agent.yaml
+
+# Check everything is consistent
+agent-catalog doctor
+```
+
+## Walkthrough: exposing agents via MCP
+
+```bash
+# Start the MCP server
+agent-catalog serve --mcp
+
+# Any MCP client can now call:
+# - catalog_list_agents
+# - catalog_get_agent
+# - catalog_search
+# - catalog_invoke
+```
+
+## Walkthrough: Python SDK
+
+```python
+from agent_catalog.client import CatalogClient
+
+client = CatalogClient()
+
+# List all production agents
+for agent in client.agents.list(env="production"):
+    print(agent.slug, agent.capability_ids())
+
+# Async
+from agent_catalog.client import AsyncCatalogClient
+client = AsyncCatalogClient()
+agents = await client.agents.list()
+```
+
+## Walkthrough: decorator API
+
+```python
+from agent_catalog import agent, capability, tool, build_manifest
+
+@agent(name="My Agent", version="1.0.0", environment="production")
+class MyAgent:
+    @capability(id="greet", description="Greets the user")
+    @tool(name="greet", description="Say hello")
+    def greet(self, name: str) -> str:
+        return f"Hello, {name}!"
+
+manifest = build_manifest(MyAgent)
+```
+
+## Commands
+
+| Command | Action |
+|---------|--------|
+| `register <path>` | Register agent from YAML |
+| `list [--env] [--output json]` | List agents |
+| `get <slug> [--output json]` | Show agent details |
+| `search [--capability] [--tool] [--surface] [--env]` | Find agents |
+| `diff <slug> [--right] [--slug2] [--right-env]` | Compare manifests |
+| `validate <path>` | Validate YAML |
+| `unregister <slug>` | Remove agent |
+| `update <slug> <path>` | Update manifest |
+| `sync <dir> [--pattern]` | Bulk register YAML files |
+| `scan <dir>` | Discover @agent-decorated Python classes |
+| `inspect <file>` | Show manifest from Python file |
+| `export-contract <slug>` | Export eval contract |
+| `security-audit [--format json]` | Security audit |
+| `graph [--format json]` | Dependency graph |
+| `serve [--port] [--mcp]` | HTTP dashboard or MCP server |
+| `run <slug> <capability>` | Invoke a capability |
+| `doctor` | Consistency check |
+
+## Storage
+
+Manifests are YAML files in `~/.config/agent-catalog/agents/`. `index.yaml` maps slugs to filenames. Set `AGENT_CATALOG_DIR` to point at a git repo — changes show up in `git diff`.
+
+```
+~/.config/agent-catalog/agents/
+├── index.yaml            # slug → filename mapping
+├── agentic-inbox.yaml    # agent manifests
+├── nexusgate.yaml
+└── trading-agent.yaml
+```
+
+## Production features
+
+- **Atomic writes.** Every write goes to a temp file then `os.replace()` — no partial writes on crash.
+- **Slug validation.** Rejects path traversal characters (/, \\, .., null).
+- **Index cache.** Repeated `index()` calls within one invocation skip re-parsing.
+- **Config validation.** Warns on type errors and unknown keys in config.yaml.
+- **Consistency check.** `doctor` command finds orphaned files and missing manifests.
+- **XSS-safe dashboard.** All agent values HTML-escaped.
+- **CORS headers.** JSON API supports cross-origin requests.
+- **Graceful shutdown.** SIGINT/SIGTERM signal handlers on HTTP server.
+- **MCP token budget.** Tool responses hard-truncated to 1500 chars.
+
+## Contributing
+
+See [CONTRIBUTING.md](CONTRIBUTING.md).
+
+## License
+
+MIT

agent_catalog_new-0.2.0/README.md

@@ -0,0 +1,173 @@
+# Agent Catalog
+
+[![CI](https://github.com/ewardGPT/agent-catalog/actions/workflows/ci.yml/badge.svg)](https://github.com/ewardGPT/agent-catalog/actions/workflows/ci.yml)
+[![PyPI](https://img.shields.io/pypi/v/agent-catalog)](https://pypi.org/project/agent-catalog/)
+[![Python](https://img.shields.io/pypi/pyversions/agent-catalog)](https://pypi.org/project/agent-catalog/)
+[![License](https://img.shields.io/github/license/ewardGPT/agent-catalog)](LICENSE)
+
+A registry for AI agents. Track every agent you own — what it does, what tools it uses, what model it runs, where it's deployed. YAML manifests stored on disk. No database. Git-ops friendly.
+
+Think `npm` for agents. One tool to list, search, diff, and invoke them all.
+
+```bash
+pip install agent-catalog
+agent-catalog register ./my-agent.yaml
+agent-catalog list
+```
+
+## Why
+
+You have agents scattered across repos, environments, and projects. Prod vs staging drift goes unnoticed. A dependency changes and nothing tells you. You grep through directories to find which agent has the capability you need.
+
+Agent Catalog fixes that. One directory is the source of truth. Every command queries it.
+
+## Features
+
+- **Registry.** Register agents from YAML manifests. List, search, get details. Filter by environment, capability, tool, surface.
+- **Diff.** Compare manifests across environments (`--left-env staging --right-env production`). Catch drift before it hits prod.
+- **Consistency.** `agent-catalog doctor` finds orphaned files and missing manifests.
+- **Security audit.** Flags MCP endpoints without auth, write capabilities without confirmation gates, idempotent tools that should not be.
+- **MCP server.** Expose the catalog as MCP tools — any MCP client can discover agents and invoke their capabilities.
+- **HTTP dashboard.** `agent-catalog serve` starts a web UI on `:8420`.
+- **Python SDK.** `from agent_catalog.client import CatalogClient` — sync or async.
+- **Decorator API.** Define agents inline with `@agent`, `@capability`, `@tool`. Manifests auto-generated from Python classes.
+- **Runtime loader.** Load an agent class from the catalog and invoke a capability from the command line or API.
+
+## Quick start
+
+```bash
+pip install agent-catalog
+
+# Register an agent
+agent-catalog register ./examples/agentic-inbox.yaml
+
+# See everything you have
+agent-catalog list
+
+# Get details
+agent-catalog get agentic-inbox
+
+# Search
+agent-catalog search --capability send_email
+
+# Validate a manifest without registering
+agent-catalog validate ./my-agent.yaml
+```
+
+## Walkthrough: managing agents across environments
+
+```bash
+# Register your staging agent
+agent-catalog register ./staging/agent.yaml --env staging
+
+# Later, register production
+agent-catalog register ./prod/agent.yaml
+
+# See the drift
+agent-catalog diff agentic-inbox --left-env staging --right-env production
+
+# Validate your changes
+agent-catalog validate ./prod/agent.yaml
+
+# Check everything is consistent
+agent-catalog doctor
+```
+
+## Walkthrough: exposing agents via MCP
+
+```bash
+# Start the MCP server
+agent-catalog serve --mcp
+
+# Any MCP client can now call:
+# - catalog_list_agents
+# - catalog_get_agent
+# - catalog_search
+# - catalog_invoke
+```
+
+## Walkthrough: Python SDK
+
+```python
+from agent_catalog.client import CatalogClient
+
+client = CatalogClient()
+
+# List all production agents
+for agent in client.agents.list(env="production"):
+    print(agent.slug, agent.capability_ids())
+
+# Async
+from agent_catalog.client import AsyncCatalogClient
+client = AsyncCatalogClient()
+agents = await client.agents.list()
+```
+
+## Walkthrough: decorator API
+
+```python
+from agent_catalog import agent, capability, tool, build_manifest
+
+@agent(name="My Agent", version="1.0.0", environment="production")
+class MyAgent:
+    @capability(id="greet", description="Greets the user")
+    @tool(name="greet", description="Say hello")
+    def greet(self, name: str) -> str:
+        return f"Hello, {name}!"
+
+manifest = build_manifest(MyAgent)
+```
+
+## Commands
+
+| Command | Action |
+|---------|--------|
+| `register <path>` | Register agent from YAML |
+| `list [--env] [--output json]` | List agents |
+| `get <slug> [--output json]` | Show agent details |
+| `search [--capability] [--tool] [--surface] [--env]` | Find agents |
+| `diff <slug> [--right] [--slug2] [--right-env]` | Compare manifests |
+| `validate <path>` | Validate YAML |
+| `unregister <slug>` | Remove agent |
+| `update <slug> <path>` | Update manifest |
+| `sync <dir> [--pattern]` | Bulk register YAML files |
+| `scan <dir>` | Discover @agent-decorated Python classes |
+| `inspect <file>` | Show manifest from Python file |
+| `export-contract <slug>` | Export eval contract |
+| `security-audit [--format json]` | Security audit |
+| `graph [--format json]` | Dependency graph |
+| `serve [--port] [--mcp]` | HTTP dashboard or MCP server |
+| `run <slug> <capability>` | Invoke a capability |
+| `doctor` | Consistency check |
+
+## Storage
+
+Manifests are YAML files in `~/.config/agent-catalog/agents/`. `index.yaml` maps slugs to filenames. Set `AGENT_CATALOG_DIR` to point at a git repo — changes show up in `git diff`.
+
+```
+~/.config/agent-catalog/agents/
+├── index.yaml            # slug → filename mapping
+├── agentic-inbox.yaml    # agent manifests
+├── nexusgate.yaml
+└── trading-agent.yaml
+```
+
+## Production features
+
+- **Atomic writes.** Every write goes to a temp file then `os.replace()` — no partial writes on crash.
+- **Slug validation.** Rejects path traversal characters (/, \\, .., null).
+- **Index cache.** Repeated `index()` calls within one invocation skip re-parsing.
+- **Config validation.** Warns on type errors and unknown keys in config.yaml.
+- **Consistency check.** `doctor` command finds orphaned files and missing manifests.
+- **XSS-safe dashboard.** All agent values HTML-escaped.
+- **CORS headers.** JSON API supports cross-origin requests.
+- **Graceful shutdown.** SIGINT/SIGTERM signal handlers on HTTP server.
+- **MCP token budget.** Tool responses hard-truncated to 1500 chars.
+
+## Contributing
+
+See [CONTRIBUTING.md](CONTRIBUTING.md).
+
+## License
+
+MIT

agent_catalog_new-0.2.0/agent_catalog/__init__.py

@@ -0,0 +1,33 @@
+"""Agent Catalog — Declarative agent registry.
+
+Discover, diff, and manage agent capabilities across environments.
+
+Decorator API (new in 0.2.0):
+
+    from agent_catalog import agent, capability, tool, interface, dependency
+    from agent_catalog.decorators import build_manifest
+"""
+
+from __future__ import annotations
+
+from agent_catalog.decorators import (
+    agent,
+    build_manifest,
+    capability,
+    dependency,
+    interface,
+    prompt_ref,
+    tool,
+)
+
+__all__ = [
+    "agent",
+    "build_manifest",
+    "capability",
+    "dependency",
+    "interface",
+    "prompt_ref",
+    "tool",
+]
+
+__version__ = "0.2.0"

agent_catalog_new-0.2.0/agent_catalog/cli/__init__.py

@@ -0,0 +1,108 @@
+"""Agent Catalog CLI — declarative agent registry.
+
+Usage:
+    agent-catalog register ./agent.yaml
+    agent-catalog list
+    agent-catalog get agentic-inbox
+    agent-catalog search --capability send_email
+    agent-catalog diff agentic-inbox --right staging
+"""
+
+from __future__ import annotations
+
+import os
+
+import typer
+from rich.console import Console
+from rich.panel import Panel
+from rich.text import Text
+
+from agent_catalog import __version__
+from agent_catalog.storage import CatalogStore
+
+# ── App ────────────────────────────────────────────────────────────────────────
+
+app = typer.Typer(
+    name="agent-catalog",
+    help="Declarative agent registry — catalog, discover, and diff agent capabilities.",
+    no_args_is_help=True,
+)
+
+# Typer adds --install-completion and --show-completion automatically.
+
+console = Console()
+
+
+# ── Version flag ────────────────────────────────────────────────────────────────
+
+
+def _version_callback(value: bool) -> None:
+    if value:
+        console.print(f"agent-catalog {__version__}")
+        raise typer.Exit()
+
+
+@app.callback()
+def _main_options(
+    version: bool = typer.Option(
+        False,
+        "--version",
+        help="Show version and exit.",
+        callback=_version_callback,
+        is_eager=True,
+    ),
+) -> None:
+    """Agent Catalog — declarative agent registry."""
+
+
+# ── Shared helpers ─────────────────────────────────────────────────────────────
+
+
+def _get_store() -> CatalogStore:
+    """Resolve the catalog store, respecting AGENT_CATALOG_DIR env var."""
+    root = os.environ.get("AGENT_CATALOG_DIR")
+    return CatalogStore(root=root) if root else CatalogStore()
+
+
+def _render_manifest(m) -> Panel:
+    """Pretty-print an agent manifest as a Rich Panel."""
+    rows: list[str] = []
+    rows.append(f"[bold]Name:[/]         {m.name}")
+    rows.append(f"[bold]Slug:[/]         {m.slug}")
+    rows.append(f"[bold]Version:[/]      {m.version}")
+    rows.append(f"[bold]Environment:[/]  {m.environment}")
+    rows.append(f"[bold]Status:[/]       {m.status}")
+    rows.append(f"[bold]Description:[/]  {m.description}")
+    if m.model:
+        rows.append(f"[bold]Model:[/]        {m.model.provider}/{m.model.name}")
+    if m.capabilities:
+        rows.append(f"[bold]Capabilities:[/] {', '.join(c.id for c in m.capabilities)}")
+    if m.tools:
+        rows.append(f"[bold]Tools:[/]        {', '.join(t.name for t in m.tools)}")
+    if m.prompt:
+        versions = ", ".join(p.version for p in m.prompt)
+        rows.append(f"[bold]Prompts:[/]      {versions}")
+    if m.dependencies:
+        deps = ", ".join(d.name for d in m.dependencies)
+        rows.append(f"[bold]Dependencies:[/] {deps}")
+    if m.eval_contract:
+        suites = ", ".join(m.eval_contract.suites)
+        rows.append(f"[bold]Eval Suites:[/]  {suites}")
+    text = "\n".join(rows)
+    return Panel(Text.from_markup(text), title=m.environment_tag())
+
+
+# ── Entry point ────────────────────────────────────────────────────────────────
+
+
+def main() -> None:
+    """Entry point for the 'agent-catalog' command."""
+    app()
+
+
+if __name__ == "__main__":
+    main()
+
+# ── Register subcommand modules ────────────────────────────────────────────────
+
+from . import aux_commands, crud, diff_commands, discover_commands  # noqa: E402,F401

agent_catalog_new-0.2.0/agent_catalog/cli/__main__.py

@@ -0,0 +1,3 @@
+from agent_catalog.cli import app
+
+app()

agent_catalog_new-0.2.0/agent_catalog/cli/aux_commands.py

@@ -0,0 +1,139 @@
+"""Auxiliary commands: security-audit, graph, serve, run, doctor."""
+
+from __future__ import annotations
+
+import json
+
+import typer
+
+from agent_catalog.cli import _get_store, app, console
+
+
+@app.command()
+def security_audit(
+    output_format: str = typer.Option("table", "--format", "-f", help="Output: table, json"),
+):
+    """Audit all registered agents for security gaps."""
+    from agent_catalog.security import audit_catalog
+
+    findings = audit_catalog(_get_store())
+    if not findings:
+        console.print("[green]No security issues found[/]")
+        return
+
+    if output_format == "json":
+        console.print(
+            json.dumps(
+                [
+                    {"severity": f.severity, "agent": f.agent, "title": f.title, "detail": f.detail}
+                    for f in findings
+                ],
+                indent=2,
+            )
+        )
+        return
+
+    from rich.table import Table
+
+    table = Table(title="Security Audit")
+    table.add_column("Severity", style="bold")
+    table.add_column("Agent")
+    table.add_column("Issue")
+    table.add_column("Detail", style="dim")
+
+    for f in findings:
+        color = {"critical": "red", "high": "yellow", "medium": "dim", "low": "dim"}[f.severity]
+        table.add_row(f"[{color}]{f.severity}[/]", f.agent, f.title, f.detail)
+
+    console.print(table)
+    crit = sum(1 for f in findings if f.severity == "critical")
+    high = sum(1 for f in findings if f.severity == "high")
+    console.print(f"[bold]Summary:[/] {len(findings)} findings ({crit} critical, {high} high)")
+
+
+@app.command()
+def graph(
+    output_format: str = typer.Option("mermaid", "--format", "-f", help="Output: mermaid, json"),
+):
+    """Show agent dependency graph."""
+    from agent_catalog.graph import build_graph, to_mermaid
+
+    store = _get_store()
+    if output_format == "json":
+        console.print_json(json.dumps(build_graph(store), default=str))
+    else:
+        console.print(to_mermaid(store))
+
+
+@app.command()
+def serve(
+    port: int = typer.Option(8420, "--port", "-p", help="HTTP port"),
+    mcp: bool = typer.Option(False, "--mcp", help="Run MCP server instead of HTTP (stdio transport)"),
+):
+    """Start the Agent Marketplace web dashboard.
+
+    By default starts an HTTP server.  Pass --mcp to run as an MCP
+    server over stdio (for use with MCP clients like Claude Code).
+    """
+    if mcp:
+        from agent_catalog.mcp_server import run_server
+
+        run_server(store=_get_store())
+    else:
+        from agent_catalog.serve import serve as run_serve
+
+        run_serve(port=port, store=_get_store())
+
+
+@app.command()
+def run(
+    slug: str = typer.Argument(..., help="Agent slug"),
+    capability: str = typer.Argument(..., help="Capability ID to invoke"),
+    params: str | None = typer.Option(None, "--params", "-p", help="JSON string of parameters"),
+):
+    """Load an agent and invoke a capability at runtime.
+
+    Uses metadata.python_module / metadata.python_class to find and
+    import the agent class, then calls the capability method.
+
+    Examples:
+        agent-catalog run my-agent greet
+        agent-catalog run my-agent greet --params '{"name": "World"}'
+    """
+    from agent_catalog.loader import invoke_capability
+
+    kwargs: dict = {}
+    if params:
+        try:
+            kwargs = json.loads(params)
+        except json.JSONDecodeError as e:
+            console.print(f"[red]\u2717[/] Invalid JSON in --params: {e}")
+            raise typer.Exit(1) from e
+
+    try:
+        result = invoke_capability(slug, capability, store=_get_store(), **kwargs)
+        console.print(result)
+    except Exception as e:
+        console.print(f"[red]\u2717[/] {e}")
+        raise typer.Exit(1) from e
+
+
+@app.command()
+def doctor():
+    """Check catalog consistency (orphaned files, missing manifests).
+
+    Scans the catalog directory and verifies every indexed agent
+    has a manifest file on disk, and every YAML file is indexed.
+    """
+    store = _get_store()
+    issues = store.check_consistency()
+
+    if not issues:
+        console.print("[green]\u2713[/] Catalog is consistent")
+        return
+
+    console.print(f"[yellow]Found {len(issues)} issue(s):[/]")
+    for issue in issues:
+        prefix = "[red]MISSING[/]" if issue.startswith("MISSING") else "[yellow]ORPHAN[/]"
+        console.print(f"  {prefix} {issue}")
+    raise typer.Exit(1)