Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add dmythro/agent-skills --skill deps-upgradegit clone --depth 1 https://github.com/dmythro/agent-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/dmythro/agent-skills/deps-upgrade)<a href="https://agentmods.dev/skills/dmythro/agent-skills/deps-upgrade"><img src="https://agentmods.dev/badge/skills/dmythro/agent-skills/deps-upgrade/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/dmythro/agent-skills/deps-upgrade"><img src="https://agentmods.dev/badge/skills/dmythro/agent-skills/deps-upgrade.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium MCP Rug Pull · line 107 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 107 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium Excessive Agency · line 248 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00134 | $0.06623 |
| Opus 5 | $0.00067 | $0.03312 |
| Sonnet 5 | $0.00027 | $0.01325 |
| Haiku 4.5 | $0.00013 | $0.00662 |
Grade A, and why
deps-upgrade scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 329 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dependency Upgrade Validation
The upgrade is the easy part; proving the project still coheres is the work. This skill takes over after versions are chosen -- typically bun update -i -r or an equivalent interactive pick -- and validates what landed: the real delta from git, peer and engine integrity, breaking changes and migrations across every version crossed, features worth adopting, and verification gates that run the project rather than reading about it. It also answers the inverse question: can a held-back package move, and why is it in package.json at all.
When to Use
- Validating a batch upgrade -- "I upgraded a lot of packages, validate everything"
- Assessing a held-back package -- "can we upgrade X?", "what blocks it?"
- Questioning a dependency -- "why do we even have X, we don't use it"
- Reading release notes -- breaking changes, migration steps, new features across the versions crossed
- Post-upgrade triage -- type, build or test failures that appeared after an update
- Pre-upgrade recon -- what a bump would cost before anyone runs it
Critical Rules
- Detect the manager first, then use only its column. A Bun project gets
bunthroughout, a pnpm projectpnpm, an npm projectnpm, a yarn projectyarn. Another manager's CLI may not be installed at all, and if it is, it resolves against its own rules and can write a second lockfile or a differently-shaped tree. Every capability this skill needs has a native form in all four managers -- see the Capability Matrix -- so there is never a reason to reach across. - Establish the delta from git, never from the prompt. "I upgraded a lot of packages" is a starting point, not an inventory. The lockfile diff is the only complete record -- it carries transitive bumps that
package.jsonnever shows. outdatedreports drift, not compatibility. A cleanbun outdatedmeans every dependency sits at the newest version its range allows; it says nothing about whether the tree still resolves, builds or runs. Never report an upgrade as validated on that basis.- Peer ranges of ecosystem packages are the ceiling, not the registry's
latest. A framework plugin pinningnext@">=16.2.6 <17.0.0"caps the framework regardless of what the registry calls latest. Read the peers of the packages that wrap the target before proposing a bump. - A peer-required package is required with zero imports.
graphqlis a mandatorypeerDependencyof[email protected]; an app that never writes a query still must declare it. Runbun whyand check every installed peer before proposing any removal (seereferences/dependency-audit.md). - Never run a migration, codemod, or version change without explicit approval. Report the exact command and what it will touch. Schema migrations additionally need an ordering and rollback story before they are worth proposing.
- Read notes for every version crossed, not just the target. A breaking change introduced in
16.9and unmentioned in16.14's notes is still a breaking change for a project coming from16.8. - Verify by running the project. Typecheck, build and tests are the evidence. A changelog that promises compatibility is not evidence.
- Report what you did not check. Skipped packages, unread notes and untested paths belong in the report; silence reads as coverage.
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 329 lines · 134 tokens per session scan A 5a4b32c0dbc7
deps-upgrade is a skill published in the GitHub repository dmythro/agent-skills (6 stars, last pushed 7d ago), licensed MIT. It adds 134 tokens to every session and 6,623 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
typescript-guardian
Focuses specifically on TypeScript type safety - any usage, unsound generics, interface design, and type narrowing. Use for a deep type-system review, not general code quality.
coding-standards
Universal coding standards, best practices, and patterns for TypeScript, JavaScript, React, and Node.js development.
claude-api
Anthropic Claude API patterns for Python and TypeScript. Covers Messages API, streaming, tool use, vision, extended thinking, batches, prompt caching, and Claude Agent SDK. Use when building applications with the Claude API or Anthropic SDKs.
typescript
TypeScript strict mode with eslint and jest.
widen-return-type
When delegating a task affected by this skill, include.
coding-standards
A set of general coding standards and practical patterns for TypeScript, JavaScript, React, and Node.js. It covers readable naming, simple designs, avoiding repetition, and delaying unnecessary features.