Publish to MCP Registry workflow (OpenOSINT/OpenOSINT)
The Publish to MCP Registry workflow from OpenOSINT/OpenOSINT, explained and optimized by Latchkey.
CI health: A - excellent
Point runs-on at Latchkey and get job timeouts, self-healing for flaky steps, and up to 58% lower cost, applied automatically.
What it does
This is the Publish to MCP Registry workflow from the OpenOSINT/OpenOSINT repository, a real project running GitHub Actions. It is shown here with attribution under its MIT license.
Below, Latchkey shows a faster, safer version produced by its optimization engine.
The workflow
name: Publish to MCP Registry
on:
release:
types: [published]
workflow_dispatch:
jobs:
publish:
runs-on: ubuntu-latest
permissions:
id-token: write
contents: read
steps:
- uses: actions/checkout@v4
- name: Install mcp-publisher
run: |
curl -L "https://github.com/modelcontextprotocol/registry/releases/latest/download/mcp-publisher_linux_amd64.tar.gz" | tar xz mcp-publisher
sudo mv mcp-publisher /usr/local/bin/
- name: Wait for PyPI to publish this version
run: |
VERSION=$(python3 -c "import json; d=json.load(open('.mcp/server.json')); print(d['packages'][0]['version'])")
echo "Waiting for openosint $VERSION on PyPI..."
for i in $(seq 1 40); do
code=$(curl -s -o /dev/null -w "%{http_code}" "https://pypi.org/pypi/openosint/$VERSION/json")
[ "$code" = "200" ] && { echo "openosint $VERSION is on PyPI"; exit 0; }
echo " try $i: got $code - retrying in 15s"
sleep 15
done
echo "Timed out waiting for openosint $VERSION on PyPI (10 min elapsed)"
exit 1
- name: Publish to MCP Registry
run: |
mcp-publisher login github-oidc
MAX_ATTEMPTS=5
DELAYS=(15 30 60 120 120)
for attempt in $(seq 1 $((MAX_ATTEMPTS + 1))); do
out=$(mcp-publisher publish .mcp/server.json 2>&1) && code=0 || code=$?
echo "$out"
if [ $code -eq 0 ]; then
echo "✅ MCP Registry publish succeeded"
exit 0
fi
# Treat an already-registered version as idempotent success.
if echo "$out" | grep -qi "duplicate version"; then
echo "✅ Version already registered in MCP Registry (idempotent)"
exit 0
fi
if [ $attempt -gt $MAX_ATTEMPTS ]; then
echo "❌ mcp-publisher publish failed after $MAX_ATTEMPTS attempts"
exit 1
fi
WAIT=${DELAYS[$((attempt-1))]}
echo "⚠️ Attempt $attempt failed. Retrying in ${WAIT}s..."
sleep "$WAIT"
done
The same workflow, on Latchkey
Removes redundant runs and caps runaway jobs. Added and changed lines are highlighted.
name: Publish to MCP Registry on: release: types: [published] workflow_dispatch: jobs: publish: timeout-minutes: 30 runs-on: latchkey-small permissions: id-token: write contents: read steps: - uses: actions/checkout@v4 - name: Install mcp-publisher run: | curl -L "https://github.com/modelcontextprotocol/registry/releases/latest/download/mcp-publisher_linux_amd64.tar.gz" | tar xz mcp-publisher sudo mv mcp-publisher /usr/local/bin/ - name: Wait for PyPI to publish this version run: | VERSION=$(python3 -c "import json; d=json.load(open('.mcp/server.json')); print(d['packages'][0]['version'])") echo "Waiting for openosint $VERSION on PyPI..." for i in $(seq 1 40); do code=$(curl -s -o /dev/null -w "%{http_code}" "https://pypi.org/pypi/openosint/$VERSION/json") [ "$code" = "200" ] && { echo "openosint $VERSION is on PyPI"; exit 0; } echo " try $i: got $code - retrying in 15s" sleep 15 done echo "Timed out waiting for openosint $VERSION on PyPI (10 min elapsed)" exit 1 - name: Publish to MCP Registry run: | mcp-publisher login github-oidc MAX_ATTEMPTS=5 DELAYS=(15 30 60 120 120) for attempt in $(seq 1 $((MAX_ATTEMPTS + 1))); do out=$(mcp-publisher publish .mcp/server.json 2>&1) && code=0 || code=$? echo "$out" if [ $code -eq 0 ]; then echo "✅ MCP Registry publish succeeded" exit 0 fi # Treat an already-registered version as idempotent success. if echo "$out" | grep -qi "duplicate version"; then echo "✅ Version already registered in MCP Registry (idempotent)" exit 0 fi if [ $attempt -gt $MAX_ATTEMPTS ]; then echo "❌ mcp-publisher publish failed after $MAX_ATTEMPTS attempts" exit 1 fi WAIT=${DELAYS[$((attempt-1))]} echo "⚠️ Attempt $attempt failed. Retrying in ${WAIT}s..." sleep "$WAIT" done
What changed
- Run on Latchkey managed runners with one line (
runs-on), which apply the fixes below automatically and self-heal transient failures. This example useslatchkey-small; pick the runner size that fits the job. - Add a job timeout so a hung step cannot burn hours of runner time.
What Latchkey heals here
This workflow has steps that commonly fail on transient issues (network, registries, flaky browsers). On Latchkey managed runners they are detected, retried, and self-healed instead of failing your build:
- Network fetches
This workflow runs 1 job per trigger. On Latchkey the same minutes cost up to 58% less than GitHub-hosted, with zero queue time.