crewAIInc/crewAI · error · SystemExit

Failed to fetch templates from GitHub: {e}

Error message

Failed to fetch templates from GitHub: {e}

What it means

Raised while paginating through the GitHub API (`/orgs/{GITHUB_ORG}/repos`) to build the template catalog for `crewai template list/add`. Any httpx.HTTPError — connection failure, timeout (15s), DNS failure, or an HTTP status error from raise_for_status() (403 rate-limit, 5xx) — triggers it. The CLI prints the red message and exits 1, chaining the original exception.

Source

Thrown at lib/cli/src/crewai_cli/remote_template/main.py:170

        console.print(panel)

    def _fetch_templates(self) -> list[dict[str, Any]]:
        """Fetch all template repos from the GitHub org."""
        templates: list[dict[str, Any]] = []
        page = 1
        while True:
            url = f"{GITHUB_API_BASE}/orgs/{GITHUB_ORG}/repos"
            params: dict[str, str | int] = {
                "per_page": 100,
                "page": page,
                "type": "public",
            }
            try:
                response = httpx.get(url, params=params, timeout=15)
                response.raise_for_status()
            except httpx.HTTPError as e:
                click.secho(f"Failed to fetch templates from GitHub: {e}", fg="red")
                raise SystemExit(1) from e

            repos = response.json()
            if not repos:
                break

            templates.extend(
                repo
                for repo in repos
                if repo["name"].startswith(TEMPLATE_PREFIX) and not repo.get("private")
            )

            page += 1

        templates.sort(key=lambda r: r["name"])
        return templates

    def _resolve_repo_name(self, name: str) -> str | None:
        """Resolve user input to a full repo name, or None if not found."""

View on GitHub (pinned to 754d7323be)

Solutions

1. Check connectivity: `curl -I https://api.github.com/orgs/crewAIInc/repos` from the same machine.
2. If behind a proxy, set HTTPS_PROXY/HTTP_PROXY (httpx honors these env vars).
3. If rate-limited (403/429 in the message), wait or set GITHUB_API_TOKEN-style auth if the CLI/deployment supports it, and reduce how often the listing runs.
4. Retry after a short wait for transient 5xx/network blips.

Example fix

# before
$ crewai template list
# Failed to fetch templates from GitHub: Client error '403 Forbidden' ...

# after
$ export HTTPS_PROXY=http://proxy.corp:8080  # or wait out the rate limit
$ crewai template list

import httpx

resp = httpx.get(
    "https://api.github.com/orgs/crewAIInc/repos",
    params={"per_page": 1, "page": 1, "type": "public"},
    timeout=15,
)
resp.raise_for_status()  # probe before running the CLI command

import time, httpx

for attempt in range(3):
    try:
        run_template_list()
        break
    except SystemExit as e:
        # transient GitHub 5xx / rate limits: back off and retry
        time.sleep(2 ** attempt)
else:
    raise

When it happens

Trigger: GET https://api.github.com/orgs/crewAIInc/repos?per_page=100&page=N with timeout=15 failing: no network, corporate proxy blocking api.github.com, GitHub secondary rate limits returning 403/429 (raise_for_status raises HTTPStatusError, an httpx.HTTPError subclass), or GitHub 5xx incidents.

Common situations: CI runners behind restrictive firewalls/proxies needing HTTP(S)_PROXY env vars; scripts hitting `crewai template list` in a loop until GitHub's unauthenticated rate limit (60 req/hr per IP) is exhausted; air-gapped environments; transient GitHub outages.

Related errors

• Failed to download template: {e}
• Unable to access repository {repo_name}: {e}
• Error. A valid pyproject.toml file is required. Check that a
• Invalid JSON response from {oauth_endpoint}
• Failed to connect to enterprise URL: {e!s}

AI-assisted analysis of crewAIInc/crewAI@754d7323be (2026-08-15). Data as JSON: /api/errors/4e76034bbdb99d16. Report an issue: GitHub.