io.github.zeynepaslierhanPublished by GitHub account zeynepaslierhanv0.7.1Updated Nov 3, 2026
Your PromptLikeEngineer prompt library as MCP tools, slash prompts and resources. No delete tool.
Connects an assistant to a PromptLikeEngineer prompt library so it can list, search, fetch, fill in and save versioned prompts.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Follow the launch instructions in the repository.
Reach your PromptLikeEngineer prompt library from inside Claude Code. Your prompts stay versioned where you saved them; Claude pulls one by name at the moment it needs it — no copy-paste, no stale duplicates in a system prompt.
From this repo's own marketplace — works today:
Inside a Claude Code session the same two steps are
/plugin marketplace add zeynepaslierhann/promptlike-plugin and
/plugin install promptlike@promptlikeengineer.
From the community marketplace (once this plugin is listed there):
Then install promptlike from @claude-community.
Or straight from a clone of this repo, handy for trying it without adding a marketplace:
Sign in once so the server can read your library (it shares the CLI session):
Without signing in the plugin still loads and its tools are listed — each call
just answers "Not logged in. Run ple login in a terminal first (install: npm i -g [email protected])."
instead of returning prompts. The suggestion hook then uses your local prompts (ple save).
The plugin starts the server with npx -y [email protected], pinned to that exact version.
Each example starts with something you do. Swap in the names of prompts from your own library.
1. Ask for a prompt by name. You type:
Claude looks the prompt up (get_prompt, after a search_prompts if the name is not exact),
fills its {{variables}} with fill_variables, and follows the result. If the name matches
several prompts, Claude shows you the candidates instead of guessing.
2. Run a prompt from the / menu. You type:
The words after the command fill the prompt's {{variables}} in the order they first appear
in it (for a prompt that mentions {{file}} and then {{focus}}, the two words above). A
variable you leave out stays as {{name}} in the text Claude receives.
3. Keep something you just worked out. You type:
The prompt tells Claude to pull the reusable instruction out of the conversation, mark what
changes between uses as {{variables}}, show you the title, description and text, and then
call save_prompt. Saving never deletes anything, and text that looks like an API key or token is refused.
Ask Claude to save a prompt, or run /mcp__plugin_promptlike_promptlike__save-this. Nothing is
ever deleted: save_prompt creates a prompt, update_prompt adds a new version and keeps the
old one, move_prompt changes where a prompt lives, and saving the same thing twice changes nothing. Plan limits
apply exactly as on the web — when you hit one, Claude tells you and nothing is saved. Content
that looks like an API key or token is refused.
On every message you send, Claude Code runs scripts/suggest.sh, which runs
ple suggest --hook (a UserPromptSubmit hook). It:
~/.promptlike when signed out). Your message is not sent anywhere;npm i -g [email protected]); without it the hook does nothing. It never blocks
your message: it answers in well under a second or stays silent.Turn it off any of these ways:
/config → PromptLikeEngineer → Suggest saved prompts → false (or run
/plugin configure promptlike@promptlikeengineer)ple config set suggest offPLE_SUGGEST=off in your environment (or in Claude Code's settings.json → env)How often a suggestion is shown is counted on your machine only (ple stats shows it);
the server only learns when Claude actually fetches a suggested prompt.
This is everything the plugin runs and every place it talks to.
Suggestion hook. Before each message Claude Code runs scripts/suggest.sh, which runs
ple suggest --hook if the ple CLI is installed (otherwise nothing happens). Matching runs
on your machine. The hook does not send your message anywhere, makes no network request
itself and never waits for one. One thing does happen in the background: if you are signed in
and the local copy of your library is missing or more than 10 minutes old, the hook starts a
detached ple suggest --refresh-cache (at most once a minute) that downloads your own prompt
list from promptlikeengineer.com with your ple login session, so the next match uses a fresh
copy. Each of the three off switches above stops the hook, and that refresh with it. When a
prompt fits, Claude gets a two-line note naming up to three of your prompt titles; like any
text in your conversation, it is sent to the model with that turn.
Files on your machine. The CLI keeps these in ~/.config/promptlike/ (or the folder named
in PROMPTLIKE_CONFIG_DIR), created with owner-only permissions where the system supports
them: config.json (your session: tokens, email and user id, written by ple login),
settings.json (the on/off switch), cache/library.json (up to 500 prompts: title,
description, tags, repo, use count and the first 600 characters of the text) and
cache/suggest-state.json (a counter and the prompts suggested in your last 20 Claude Code
sessions). Signed out, the hook reads ~/.promptlike instead.
MCP server. Claude Code starts it with npx -y [email protected]; the first time, npm
downloads that exact version from the npm registry. It runs on your machine and talks only to
https://promptlikeengineer.com (or the address in PROMPTLIKE_API_URL, if you set one),
authenticating with the session ple login stored and refreshing it through the same host
when it expires. It has no other network code. What it sends:
list_prompts, search_prompts, get_prompt, fill_variables) ask your account
for your prompts; Claude Code also asks for the list that fills the / menu and the @
suggestions (up to 500 prompts, kept for 30 seconds);/ menu or as an @ resource — the server
records a use event: which prompt and version, which of those three ways it was reached,
whether it was offered as a suggestion, and the MCP client's name and version (for example
claude-code/2.1.179);suggest_prompts sends the task description Claude gives it (up to 2,000 characters) so the
server can match it against your library; the endpoint does not log or store that text. The
hook above never does this;save_prompt and update_prompt send the text you asked to save;submit_feedback runs only after Claude shows you the sentence and you say yes. It sends that one short sentence, a type (bug, missing, idea or prompt), the title or id of the prompt it is about if any, and the server version, OS and CPU, Node version and tool name. It works without a session, never carries a prompt body, and refuses a message that looks like a key or token.Limits on what it does. Nothing is deleted: the server has no delete tool, saving the same text twice changes nothing, and content that looks like a secret (API keys, tokens, private keys) is refused — the error names the kind, never the value. Prompts you save live in your PromptLikeEngineer account; how the service handles them is in its Privacy Policy and Terms.
Point at another deployment with PROMPTLIKE_API_URL (the same environment variable the
CLI honors).
Questions and bug reports: open an issue at
https://github.com/zeynepaslierhann/promptlike-plugin/issues or write to Zeynep Aslı Erhan ·
[email protected]. Please report a security problem by email rather than in a public
issue. This repository holds only the plugin files; the MCP server is the npm package
promptlike-mcp and the CLI you sign in with
is promptlike, both MIT licensed.
MIT licensed.
Source: README.md at commit 20e85b0
List prompts
List the user's saved prompts from their PromptLikeEngineer library. Optionally filter by tag, repo (`project`) or `folder`. Returns id, title, description, repo, folder, updated_at and the {{variable}} names each prompt expects. Paged: `total` is the full count, and when `has_more` is true call again with `page` + 1.
tagpagelimitfolderprojectSearch prompts
Search the user's prompt library by title and description. Returns the same (paged) shape as list_prompts.
pagelimitqueryrequiredGet prompt content
Fetch the full content of one prompt by name, title, id or short-id. Optionally pin an exact version number or a deployment label (e.g. production). Returns the prompt text plus the {{variable}} names it expects — if there are variables, either fill them yourself or call fill_variables.
vianamerequiredlabelprojectversionFill prompt variables
Fetch a prompt and substitute its {{variable}} placeholders with the given values, returning ready-to-use text. Empty or unset ones stay as {{name}} in the output (missing_variables); names you passed that the prompt does not have are ignored and listed in unknown_variables, so check it before relying on the text.
namerequiredlabelprojectversionvariablesrequiredSuggest saved prompts
Call once at the start of a non-trivial task to check the user's saved prompts. Returns up to `limit` prompts from their own library that match the task (title/description match, weighted by past use). If one fits, fetch it with get_prompt (via: 'suggestion'). An empty list means nothing fits — carry on normally. The task text is not stored.
taskrequiredlimitSave a new prompt
Save a new prompt to the user's PromptLikeEngineer library (use when the user asks to keep/save a prompt, or via the save-this prompt). Mark reusable parts as {{variables}}; never include secrets. Saving the same title + content again is a no-op. If the title already exists with different content, call update_prompt instead. Plan limits apply (the error says so).
tagstitlerequiredfoldercontentrequiredprojectdescriptionUpdate a prompt (new version, description, tags)
Change an existing prompt. `name` must be the exact title or the id. Only the fields you pass are touched; a field you leave out is NOT changed, a field you pass OVERWRITES. `content` (the full new text) is saved as a NEW version: the old version is kept and nothing is deleted; content equal to the current version creates no version. bump: patch (default), minor or major. `description` (one line; '' clears it) overwrites the prompt's description. `tags` ADDS tags: the ones the prompt lacks are added; tags already on it, and tags you do not list, are left as they are. It never removes a tag ([] changes nothing; remove a tag in the web app). Description and tags are not part of a version: changing only them never creates one, and the previous description is returned so it can be restored. Pass at least one of content, description, tags; passing values that already match changes nothing. This tool never deletes a prompt, a version or a tag.
bumpnamerequiredtagscontentprojectchangelogdescriptionMove a prompt to another repo or folder
Move ONE prompt to another repo and/or folder. `name` must be the exact title or the id. Its versions, deployment labels, stars, usage and its /p/ short link stay with it; tags are matched by name in the target repo (created if missing). `to_project` = target repo (omit to stay in the same repo; then pass `folder`). `folder` = target folder like reviews/backend (omit: kept in the same repo, repo root in another; "" = the repo root). Refused, with nothing moved, when the target repo already has a prompt with that title, or when the move changes who can see the prompt (public <-> private repo): ask the user, and only if they agree call again with confirm_visibility: true. Moving to where it already is changes nothing. Needs editor or owner on both repos. This tool never deletes a prompt, a version or a tag.
namerequiredfolderprojectto_projectconfirm_visibilitySend feedback to the PromptLikeEngineer maker
Call this when a PromptLikeEngineer tool is missing, broken, confusing or returned a wrong result, or when a saved prompt did not work. One short factual sentence is enough. Never include prompt bodies, secrets or personal data. Works without a signed-in session. A call with a key or token in the message is refused. Errors come back as JSON with `code` and `retryable`; do not retry in a loop.
typeemailpromptmessagerequired90% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
claude plugin marketplace add zeynepaslierhann/promptlike-pluginclaude plugin install promptlike@promptlikeengineer/plugin marketplace add anthropics/claude-plugins-communitygit clone https://github.com/zeynepaslierhann/promptlike-pluginclaude --plugin-dir ./promptlike-pluginnpm i -g [email protected] && ple loginRun my code-review prompt on the changes in this branch./mcp__plugin_promptlike_promptlike__code-review src/auth.ts security/mcp__plugin_promptlike_promptlike__save-this