Running Wintermolt
Use when building, running, launching, smoke-testing, or verifying the wintermolt binary — covers zig build, single-shot -e mode, driving the REPL, backend requirements (Ollama), and crash diagnosis.
Install
npx skills add lupin4/wintermolt --skill running-wintermoltgit clone --depth 1 https://github.com/lupin4/wintermolt.git /tmp/wintermolt
mkdir -p ~/.claude/skills
cp -r /tmp/wintermolt/.claude/skills/running-wintermolt ~/.claude/skills/running-wintermoltPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Running Wintermolt Wintermolt is a single Zig binary: an agentic AI CLI (the MIT/open-source lite version of Wintermute, an OpenClaw-like assistant). Multi-backend (Ollama, Claude, OpenAI-compat, local Metal "kernel"), with tools, MCP, and skills. ## Build ```bash zig build # → zig-out/bin/wintermolt (silent on success) ``` Needs system libcurl + sqlite3 (preinstalled on macOS). Zig 0.16.0 (via zigup at ~/.local/share/zigup/); 0.15.2 no longer builds wintermolt. Cross-compile: `zig build -Dtarget=aarch64-linux-gnu` (Jetson/Pi) or `x86_64-linux-gnu`. A PUBLISHED binary is built differently — never a plain `zig build`, which bakes in native CPU features. See the rule at build.zig:48: ```bash zig build -Doptimize=ReleaseFast -Dcpu=baseline -Dstrip=true # aarch64 zig build -Doptimize=ReleaseFast -Dcpu=x86_64_v3 -Dstrip=true # x86_64 ``` `-Dcpu=` is the knob, not `-Dtarget=`: naming a target makes Zig treat the build as a cross-compile and stop finding libcurl/libsqlite3. On aarch64, verify before publishing — this must print 0: ```bash objdump -d zig-out/bin/wintermolt \ | grep -cE '\b(ptrue|movprfx|whilelo|ld1d|st1d|cntd)\b' ``` ## Run + verify (the smoke test) **Single-shot** — exercises the full agent loop, exits when done: ```bash ./zig-out/bin/wintermolt -e "Reply with exactly: WINTERMOLT OK" ``` Expected: `WINTERMOLT OK` on stdout. Two benign stderr lines always appear
Details
- Skill name
running-wintermolt- Repository
- lupin4/wintermolt
- Path
.claude/skills/running-wintermolt/SKILL.md- License
- MIT
- Repo updated
- 14 November 2026
FAQ
What does the Running Wintermolt skill do?
Use when building, running, launching, smoke-testing, or verifying the wintermolt binary — covers zig build, single-shot -e mode, driving the REPL, backend requirements (Ollama), and crash diagnosis.
How do I install the Running Wintermolt skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add lupin4/wintermolt --skill running-wintermolt. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Running Wintermolt popular?
It lives in lupin4/wintermolt, which has 49 GitHub stars and 5 forks.
Related skills
Auditar
100adriannoes/awesome-agentic-ai
Use to audit the conformance of the SDD pipeline — runs the structural validator (frontmatter, links, specs) and checks what requires judgment (AC→test→commit traceability, DoD, orphan specs, live docs up to date). Reports violations with the file. Trigger with /auditar.
Sandbox Self Test
81TomGranot/hubspot-admin-skills
Verify the entire toolkit against a disposable HubSpot developer test account or sandbox that you bring: seed synthetic fixtures, run every scripted skill's read path, exercise end-to-end and API round-trip cases, produce a graded report, and tear everything down. Hard-refuses to run against production.
M365 Entra Attack
100adriannoes/awesome-agentic-ai
Microsoft 365 / Entra ID red-team attack chain — current 2026 reality. AADSTS code reference, user enumeration vectors (with hardening status), Smart Lockout math, Conditional Access bypass options, ROPC + SAML SSO browser flow, Burp/Playwright templates. Built from authorized red-team work where ROPC spray surfaced pre-existing lockouts and CA-blocked credentials, plus real-time external attacker activity correlation. Use for any M365/Entra credential attack, password spray, user enumeration, CA-bypass exploration, or active-attacker-detection scenario.
Device Class Authoring
44netdisco/snmp-info
Use when adding or changing an SNMP::Info device class (lib/SNMP/Info/Layer*/Vendor.pm), its detection in device_type(), its tests or its POD, or when choosing between a standard MIB and a vendor MIB for a method.
Mobile Native
100adriannoes/awesome-agentic-ai
Make a web app feel native on a phone — the small CSS and meta-tag fixes that separate "a website in a browser" from something that feels installed. Covers sticky hover states, tap highlight flashes, the 100vh bug, inputs that zoom the page, laggy taps, pull-to-refresh hijacking scroll, content under the notch, long-press selecting button text, carousels that scroll the wrong way, mismatched status bars, and the rule that you test on real hardware. Use when a web app is being built for or reviewed on mobile, when something "works in Chrome but feels wrong on my phone", when building a PWA, a b
Cleanup Properties
81TomGranot/hubspot-admin-skills
Archive or delete unused custom properties across all HubSpot object types (contacts, companies, deals). Identifies Salesforce sync properties, test/temp properties, and obsolete form fields.
Browse the most popular Claude skills, all testing & qa skills, or write your own with the skills generator.