Writing

v1.1.0

Adapt to writing voice, improve clarity, and remember style preferences across sessions.

6· 5.1k·55 current·59 all-time
byIván@ivangdavila
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description (adapt to writing voice, remember preferences) matches the instructions and required artifacts. The skill only asks to create and use a ~/writing/ directory for tiered memory and per-project notes, which is appropriate for a persistent writing-preferences feature.
Instruction Scope
SKILL.md is explicit about operations: read/write ~/writing/memory.md, load project files, prompt user before saving, and offer commands to view/clear memory. These actions are within the domain of a memory-backed writing helper. Note: detection triggers describe automatic activation when certain user phrases appear; that can cause the agent to enter 'writing mode' without an extra explicit prompt, so the activation policy should be clear to the user (the skill itself instructs to get explicit consent before saving).
Install Mechanism
No install spec and no code files — instruction-only. This is the lowest-risk install approach and matches the declared architecture.
Credentials
The skill requires no environment variables, binaries, or external credentials. It only references local file paths under ~/writing/, which is proportionate to remembering writing preferences.
Persistence & Privilege
always:false (not forced into every agent run). The skill is allowed to be invoked autonomously (platform default) and will persist preferences locally across sessions. This persistence is coherent with the stated purpose, but users should be aware that local files will be created and updated over time. The skill documents asking for explicit consent before saving, which mitigates the privacy risk.
Assessment
This skill appears to do what it says: it stores and uses writing preferences under ~/writing/ and promises no network access. Before installing, consider: 1) Confirm you are comfortable with the skill creating and updating files in your home directory (~/writing/). 2) The skill can activate automatically on trigger phrases — verify activation preferences during setup and only enable proactive behavior if you want it. 3) Do not put secrets or credentials into memory.md or project files; treat these files as persistent user data. 4) Periodically inspect ~/writing/memory.md (and project files) and use the provided "Forget my style"/clear commands if you want to remove stored preferences. 5) Because this is instruction-only, its actual behavior depends on the agent implementation executing those instructions — if you use other skills that have network access, be mindful of combined behaviors (e.g., a separate skill that uploads files could read these local files). If you want extra caution, test the skill with non-sensitive examples and keep backups of any important content.

Like a lobster shell, security has layers — review code before you run it.

latestvk97e5ykptvsd9j5vdx0p5y06rs81s6vt

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Runtime requirements

✍️ Clawdis
OSLinux · macOS · Windows

Comments