Dependency Update Bot logo

Dependency Update Bot

Organization
Varnan-Tech
dependency-update-bot

Scans your project for outdated npm, pip, Cargo, Go, or Ruby packages. Runs a CVE security audit. Fetches changelogs, summarizes breaking changes with Gemini, and opens one PR per risk group (patch, minor, major). Includes Diagnosis Mode for install conflicts. Use when asked to update dependencies, check for outdated packages, open dependency PRs, scan for package updates, audit for CVEs, or flag breaking changes in upgrades. Trigger when a user says "check for outdated packages", "update my dependencies", "open PRs for dependency updates", "scan for CVEs", or "which packages need upgrading".

Overview

PublisherVarnan-Tech
Repositoryopendirectory
Skill namedependency-update-bot
Stars
674
Forks
74
Bundled files
2
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 2 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by Varnan-Tech on GitHub. Read the source before you install it.

Installation

Install the Dependency Update Bot AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/Varnan-Tech/opendirectory/tree/main/skills/dependency-update-bot
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/Varnan-Tech/opendirectory.git /tmp/opendirectory
mkdir -p .claude/skills
cp -r /tmp/opendirectory/skills/dependency-update-bot .claude/skills/dependency-update-bot
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Dependency Update Bot in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Dependency Update Bot on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Dependency Update Bot is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Dependency Update Bot

Scan for outdated packages. Run a security audit. Fetch changelogs. Summarize breaking changes. Open one PR per risk group.


Critical rule: Only update packages that the package manager's outdated command actually reports. Never guess or invent version numbers. If a changelog cannot be fetched, note the gap rather than inventing content.


Step 1: Setup Check

bash
echo "GEMINI_API_KEY: ${GEMINI_API_KEY:+set}"
echo "GITHUB_TOKEN: ${GITHUB_TOKEN:-not set, changelog fetching rate-limited to 60/hour}"
gh auth status 2>/dev/null | head -1 || echo "gh: not authenticated"

If GEMINI_API_KEY is missing: Stop. Tell the user: "GEMINI_API_KEY is required. Get it at aistudio.google.com. Add it to your .env file."

If gh is not authenticated: Stop. Tell the user: "GitHub CLI must be authenticated. Run: gh auth login"

Detect package manager(s):

bash
ls package.json 2>/dev/null && echo "npm"
ls requirements.txt pyproject.toml 2>/dev/null && echo "pip"
ls Cargo.toml 2>/dev/null && echo "cargo"
ls go.mod 2>/dev/null && echo "go"
ls Gemfile 2>/dev/null && echo "ruby"

If multiple are found, ask: "Found [list]. Which should I scan? (all / npm / pip / cargo / go / ruby)"


Step 2: Detect Outdated Packages

npm:

bash
npm outdated --json --long 2>/dev/null | python3 -c "
import sys, json
data = json.load(sys.stdin)
for name, info in data.items():
    print(json.dumps({'name': name, 'current': info.get('current','?'), 'latest': info.get('latest','?'), 'dep_type': info.get('type','dependencies')}))
"

pip:

bash
pip list --outdated --format=json 2>/dev/null | python3 -c "
import sys, json
for p in json.load(sys.stdin):
    print(json.dumps({'name': p['name'], 'current': p['version'], 'latest': p['latest_version']}))
"

Cargo (Rust):

bash
cargo outdated --format json 2>/dev/null || \
  cargo outdated 2>/dev/null | grep -v "^---" | tail -n +3 | head -30
# If cargo-outdated not installed: cargo install cargo-outdated

Go modules:

bash
go list -u -m -json all 2>/dev/null | python3 -c "
import sys, json
decoder = json.JSONDecoder()
buf = sys.stdin.read()
pos = 0
while pos < len(buf):
    try:
        obj, idx = decoder.raw_decode(buf, pos)
        if obj.get('Update'):
            print(json.dumps({'name': obj['Path'], 'current': obj['Version'], 'latest': obj['Update']['Version']}))
        pos += idx
    except: break
"

Ruby (Bundler):

bash
bundle outdated --parseable 2>/dev/null | python3 -c "
import sys
for line in sys.stdin:
    parts = line.strip().split()
    if len(parts) >= 4:
        print('{\"name\":\"' + parts[0] + '\",\"current\":\"' + parts[3].strip('()') + '\",\"latest\":\"' + parts[1] + '\"}')
"

If all return empty: "All packages are up to date." Stop.

State count before proceeding: "Found X outdated packages."


Step 3: Classify by Risk Level

Parse version bump (current → latest):

  • MAJOR: first digit changed (1.x.x → 2.x.x)
  • MINOR: second digit changed (1.2.x → 1.3.x)
  • PATCH: third digit changed (1.2.3 → 1.2.4)
bash
python3 -c "
def classify(current, latest):
    try:
        c = [int(x) for x in current.lstrip('v').split('.')[:3]]
        l = [int(x) for x in latest.lstrip('v').split('.')[:3]]
        if l[0] > c[0]: return 'major'
        if len(l) > 1 and len(c) > 1 and l[1] > c[1]: return 'minor'
        return 'patch'
    except: return 'unknown'
"

State the breakdown: "Patch: X packages. Minor: Y packages. Major: Z packages."


Step 4: Security Audit

Run a CVE scan before creating any PRs. This determines urgency.

npm:

bash
npm audit --json 2>/dev/null | python3 -c "
import sys, json
d = json.load(sys.stdin)
vulns = d.get('vulnerabilities', {})
for pkg, info in vulns.items():
    sev = info.get('severity', 'unknown')
    via = [v.get('title','') for v in info.get('via',[]) if isinstance(v, dict)]
    print(f'  [{sev.upper()}] {pkg}: {via[0] if via else \"see npm audit\"}')
" 2>/dev/null || echo "No vulnerabilities found or npm audit not available"

pip:

bash
pip-audit --format=json 2>/dev/null | python3 -c "
import sys, json
for vuln in json.load(sys.stdin):
    print(f'  [{vuln.get(\"aliases\",[\"\"])[0]}] {vuln[\"name\"]} {vuln[\"version\"]}: {vuln[\"description\"][:80]}')
" 2>/dev/null || echo "pip-audit not installed. Run: pip install pip-audit"

Cargo:

bash
cargo audit 2>/dev/null | grep -E "^(ID|Package|Severity|URL)" | head -30 \
  || echo "cargo-audit not installed. Run: cargo install cargo-audit"

Escalation rule: If a PATCH or MINOR update has a Critical or High CVE, promote it to MAJOR priority: it gets its own PR and the CVE details go in the PR body.

Report security findings before proceeding:

Security audit: [N] vulnerabilities found
  [CRITICAL] lodash 4.17.19: Prototype Pollution (CVE-2021-23337)
  [HIGH] axios 0.21.1: Server-Side Request Forgery (CVE-2021-3749)

If no vulnerabilities: "Security audit: clean."


Step 5: Fetch Changelogs

For each package, try sources in order. Stop at first that returns content.

Source 1: GitHub Releases API

Get repo URL from registry:

bash
# npm
curl -s "https://registry.npmjs.org/{PACKAGE}/latest" \
  | python3 -c "import sys,json; d=json.load(sys.stdin); r=d.get('repository',{}); print(r.get('url','') if isinstance(r,dict) else str(r))"

# pip
curl -s "https://pypi.org/pypi/{PACKAGE}/json" \
  | python3 -c "import sys,json; d=json.load(sys.stdin); print(d.get('info',{}).get('home_page','') or d.get('info',{}).get('project_urls',{}).get('Source',''))"

Fetch last 5 releases:

bash
AUTH_HEADER=""
[ -n "$GITHUB_TOKEN" ] && AUTH_HEADER="-H \"Authorization: Bearer $GITHUB_TOKEN\""
curl -s $AUTH_HEADER \
  "https://api.github.com/repos/{OWNER}/{REPO}/releases?per_page=5" \
  | python3 -c "import sys,json; [print(json.dumps({'tag':r.get('tag_name',''),'body':r.get('body','')[:1500]})) for r in json.load(sys.stdin)]"

Keep releases between current and latest version only.

Source 2: npm registry README (fallback)

bash
curl -s "https://registry.npmjs.org/{PACKAGE}" \
  | python3 -c "import sys,json; print(json.load(sys.stdin).get('readme','')[:3000])"

Source 3: PyPI description (last resort for pip)

bash
curl -s "https://pypi.org/pypi/{PACKAGE}/json" \
  | python3 -c "import sys,json; print(json.load(sys.stdin).get('info',{}).get('description','')[:2000])"

If no source returns content: note "No changelog found" and continue.


Step 6: Summarize with Gemini

One request per risk group. Include security findings for any CVE-affected packages:

bash
cat > /tmp/deps-summary-request.json << 'ENDJSON'
{
  "system_instruction": {
    "parts": [{
      "text": "You are a developer writing a GitHub PR description for a dependency update. Given a list of packages being updated and their raw changelog content, write a concise PR body in Markdown. Rules: For each package, list only what changed between the OLD version and the NEW version. Use bullet points. Flag breaking changes with a BREAKING prefix. Flag CVE fixes with a SECURITY prefix and include the CVE ID. Keep each package section to 3-5 bullets maximum. If no changelog was found for a package, write 'No changelog available.' Do not use em dashes. Do not use these words: seamless, robust, leverage, transform, innovative. Output only the Markdown PR body, no commentary."
    }]
  },
  "contents": [{
    "parts": [{
      "text": "PACKAGES_AND_CHANGELOGS_HERE"
    }]
  }],
  "generationConfig": {
    "temperature": 0.2,
    "maxOutputTokens": 2048
  }
}
ENDJSON

curl -s -X POST \
  "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=$GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d @/tmp/deps-summary-request.json \
  | python3 -c "import sys,json; d=json.load(sys.stdin); print(d['candidates'][0]['content']['parts'][0]['text'])"

Step 7: Create PRs

One PR per non-empty risk group. One PR per package for major updates (individual review required).

1. Create branch:

bash
BRANCH="deps/{RISK}-updates-$(date +%Y%m%d)"
git checkout -b "$BRANCH"

2. Update package file:

npm:

bash
npm install {package}@{latest_version} --save-exact
# devDependencies:
npm install {package}@{latest_version} --save-dev --save-exact

pip:

bash
python3 -c "
import re, sys
pkg, version, filename = sys.argv[1], sys.argv[2], sys.argv[3]
with open(filename) as f: content = f.read()
pattern = rf'^{re.escape(pkg)}[>=<!\s].*$'
new_content = re.sub(pattern, f'{pkg}=={version}', content, flags=re.MULTILINE|re.IGNORECASE)
if new_content == content: new_content = content + f'\n{pkg}=={version}'
open(filename, 'w').write(new_content)
" "{PACKAGE}" "{LATEST}" "requirements.txt"

Cargo:

bash
# Edit Cargo.toml version field for the package, then:
cargo update {package}

Go:

bash
go get {module}@{latest_version}
go mod tidy

Ruby:

bash
bundle update {gem_name}

3. Commit:

bash
git add -A
git commit -m "chore(deps): update {RISK} dependencies $(date +%Y-%m-%d)"

4. Create PR:

bash
cat > /tmp/dep-pr-body-{RISK}.md << 'ENDMD'
PR_BODY_FROM_GEMINI
ENDMD

gh pr create \
  --title "chore(deps): update {RISK} dependencies" \
  --body-file /tmp/dep-pr-body-{RISK}.md \
  --label "dependencies" \
  --base main

Major updates get label dependencies,breaking-change. CVE-fixing updates get label dependencies,security.

After each PR, return to main: git checkout main


Step 8: Diagnosis Mode

Trigger: If any package install command fails mid-run, enter Diagnosis Mode instead of stopping.

Detect the failure type:

Error patternLikely causeSuggested fix
peer dep conflictPeer dependency incompatibilityShow conflicting pair, suggest --legacy-peer-deps flag or downgrade
ERESOLVEnpm resolution conflictRun npm install --legacy-peer-deps for the affected package only
version not foundVersion does not exist in registryCheck registry with npm view {pkg} versions
python requiresPython version incompatibilityNote required Python version, skip package
cargo E0463Rust edition incompatibilityFlag for manual review

Present a diagnosis summary:

Install failed for {package}: {error type}
Likely cause: {explanation}
Suggested fix: {specific command or action}
Remaining packages: proceeding with {N} that succeeded.

Do not stop the entire run when one package fails. Continue with packages that succeed.


Step 9: Output Summary

## Dependency Update Summary: [YYYY-MM-DD]

### Security
[CRITICAL] lodash: CVE-2021-23337 fixed in 4.17.21: PR #42
[HIGH] axios: CVE-2021-3749 fixed in 0.21.4: PR #42

| Risk Level | Packages | PR |
|------------|----------|-----|
| Patch | lodash 4.17.19→4.17.21, axios 0.21.1→0.21.4 | #42 |
| Minor | express 4.17.1→4.18.2 | #43 |
| Major | react 17.0.2→18.2.0 | #44 |

PRs opened: 3

Packages with no changelog: some-obscure-pkg (no GitHub repo in registry)
Install failures: none

Next action: Review major update PRs individually before merging.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Dependency Update Bot AI skill do?

Scans your project for outdated npm, pip, Cargo, Go, or Ruby packages. Runs a CVE security audit. Fetches changelogs, summarizes breaking changes with Gemini, and opens one PR per risk group (patch, minor, major). Includes Diagnosis Mode for install conflicts. Use when asked to update dependencies, check for outdated packages, open dependency PRs, scan for package updates, audit for CVEs, or flag breaking changes in upgrades. Trigger when a user says "check for outdated packages", "update my dependencies", "open PRs for dependency updates", "scan for CVEs", or "which packages need upgrading".

Why use Dependency Update Bot on TypingMind?

Because you install it once and use it with any model. Dependency Update Bot is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Dependency Update Bot in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Varnan-Tech/opendirectory/tree/main/skills/dependency-update-bot. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Dependency Update Bot?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Dependency Update Bot?

As many as you like. As long as a model supports skills, you can use Dependency Update Bot with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Dependency Update Bot AI skill free?

Yes. It is published on GitHub by Varnan-Tech under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇