Run an openheard feedback board through its MCP server - connect it, set up a workspace from a website, install the widget in this codebase, triage and dedupe feedback, reply, ship and announce…
Install the "openheard" agent skill from https://github.com/Heilonng23/openheard/tree/main/skills/openheard into .claude/skills/openheard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "openheard", then confirm the skill loads.
Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Type this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
skills CLI
$ npx skills add Heilonng23/openheard --skill openheard -a codex
Project install goes to .agents/skills/; add -g for ~/.codex/skills/.
Install the "openheard" agent skill from https://github.com/Heilonng23/openheard/tree/main/skills/openheard into .agents/skills/openheard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "openheard", then confirm the skill loads.
Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
skills CLI
$ npx skills add Heilonng23/openheard --skill openheard -a cursor
Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.
Install the "openheard" agent skill from https://github.com/Heilonng23/openheard/tree/main/skills/openheard into .cursor/skills/openheard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "openheard", then confirm the skill loads.
Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
skills CLI
$ npx skills add Heilonng23/openheard --skill openheard -a gemini-cli
Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.
Install the "openheard" agent skill from https://github.com/Heilonng23/openheard/tree/main/skills/openheard into .gemini/skills/openheard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "openheard", then confirm the skill loads.
Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
GitHub CLI
$ gh skill install Heilonng23/openheard openheard
Installs for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
skills CLI
$ npx skills add Heilonng23/openheard --skill openheard -a github-copilot
Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.
Install the "openheard" agent skill from https://github.com/Heilonng23/openheard/tree/main/skills/openheard into .github/skills/openheard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "openheard", then confirm the skill loads.
GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
skills CLI
$ npx skills add Heilonng23/openheard --skill openheard -a opencode
OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
Install the "openheard" agent skill from https://github.com/Heilonng23/openheard/tree/main/skills/openheard into .opencode/skills/openheard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "openheard", then confirm the skill loads.
OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
Skill name
openheard
GitHub stars
175
Token cost
~1.5k tokens
SKILL.md length
804 words
Files
4
Skills in repo
1
Repo updated
First seen
Licence
AGPL-3.0
At a glance
Run an openheard feedback board through its MCP server - connect it, set up a workspace from a website, install the widget in this codebase, triage and dedupe feedback, reply, ship and announce…
Works in 5 steps: Connect → Set up a new workspace → Weekly triage → …
The user mentions openheard
SKILL.md covers Rules, 1. Connect, 2. Set up a new workspace and 3. Weekly triage, plus 3 more sections
Openheard is an agent skill from Heilonng23/openheard. Run an openheard feedback board through its MCP server - connect it, set up a workspace from a website, install the widget in this codebase, triage and dedupe feedback, reply, ship and announce changelog entries, keep the help center current. Use when the user mentions openheard, feedback triage, their roadmap, changelog, help center or feedback widget.
Its SKILL.md is about 1.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 4 other files (for example `reference/install.md`, `reference/tools.md` and `reference/widget-install.md`).
It sits in Sales & Support, covering Help center and FAQ content and Changelog and release notes. It works with Model Context Protocol. The repository describes itself as: open source canny / featurebase alternative. The licence is AGPL-3.0.
When your agent uses it
The user mentions openheard
Feedback triage
Feedback widget
Example prompts
“/openheard”
Requirements
A credential in OPENHEARD_KEY
Workflow steps
5 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 4376828. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Shell commands in SKILL.md call:
claude
From the folder's file list and the shell code blocks in SKILL.md.
Network
Hosts in commands or code, which the agent is likely to contact:
openheard.com
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names these keys or tokens, usually read from environment variables:
OPENHEARD_KEY
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Openheard loads about 1.5k tokens when it runs. Until then it costs about 91 tokens; SKILL.md has 804 words of instructions outside code blocks.
Always· name and description, kept in context so the agent knows when to use it
~91
When it runs· the whole SKILL.md, loaded when a task matches
~1.5k
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
Download SKILL.mdSave it as .claude/skills/openheard/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.
name
openheard
description
Run an openheard feedback board through its MCP server - connect it, set up a workspace from a website, install the widget in this codebase, triage and dedupe feedback, reply, ship and announce changelog entries, keep the help center current. Use when the user mentions openheard, feedback triage, their roadmap, changelog, help center or feedback widget.
openheard
openheard is a feedback board with voting, a roadmap, a changelog, a help center and an embeddable widget. Its MCP server (/api/mcp) exposes every admin action as a tool. Every write runs the dashboard's own code, so status changes and changelog publishes email real people.
Rules
Ask before anything people will see or get emailed: set_status (emails followers), publish_changelog (emails voters and subscribers), add_comment (public), invite_member.
Destructive tools (delete_*, merge_posts, disconnect) need confirm: true. Show what goes, get a yes, then set it.
Run list_workspaces first. One key reaches all the user's workspaces: if there is more than one, ask which to use and pass workspace on every call. Never guess a slug. Every result names the workspace it acted on.
Boards, statuses and tags accept names. When a call fails, the error lists what exists; use that list rather than guessing again.
Link things in replies: every result carries a url.
1. Connect
If the openheard tools are missing, the server is not connected:
The user makes a key: Settings > API keys. One key reaches all their workspaces and can create new ones. Limit to this workspace is for a key shared with a teammate or a script.
Self-hosted: use the install's own origin. Cursor and Claude Desktop: see reference/install.md.
Check with list_workspaces.
Never paste the key into a committed file. For .mcp.json use ${OPENHEARD_KEY} from the environment.
2. Set up a new workspace
match_website with the product site. Show the name, accent, theme and logo.
No workspace for this product yet: create_workspace (name, slug of 5+ characters, website). The website's brand is applied on creation. Otherwise use apply_branding with website, or with the fields the user kept.
list_boards / list_statuses. Propose 2 to 4 boards that fit the product (Feature requests exists already; add for example Bugs or Integrations) and any extra status (kind decides behaviour: done = shipped, closed = declined). Create them after a yes.
configure_widget: theme auto, accent brand, launcher and position to taste, allowed_sites with the production origin (plus http://localhost:<port> while developing).
Detect the framework from package.json (next, nuxt, vue, react/vite) or plain HTML.
get_widget_snippet with that framework.
Add it once to the root layout file named in the result. Edit the existing file; never replace it. Show the diff.
Finish with the board URL, the dashboard URL and what the user still has to do (deploy, invite the team with invite_member, connect Slack with connect_slack then send_test).
Show full SKILL.md (384 more words)Show less
3. Weekly triage
summarize_feedback. Report: new this week, top voted, rising, stale planned or in-progress items, and posts nobody has answered (needsReply).
Duplicates: find_duplicates with the post_id of each new or rising post. Only suggest a merge when the titles really mean the same thing. Merge into the post with more votes (merge_posts from=<duplicate> into=<keeper> confirm=true after a yes).
Statuses: propose one per open post in a table (planned, closed with a one-line reason, or leave open). Apply with bulk_set_status. Pass note when the reason matters to voters.
Replies: draft short public replies for needsReply posts. Thank the author, answer or say what happens next, and link a help article if search_help_articles finds one. Post them with add_comment. Context for the team goes in add_internal_note.
Stale items: ask whether each stale planned post is still planned. Update the eta with update_post, or move it back.
4. Ship and announce
Find what shipped: list_posts with status planned or progress, plus find_duplicates with the feature description as text. Confirm the list with the user.
draft_changelog: a user-facing title, a short markdown body (what changed, why it matters, how to use it), version if they tag releases, and post_ids.
Before publishing, say who hears about it. publish_changelog moves every linked post to done and emails its voters and followers once, and emails confirmed changelog subscribers. Editing afterwards (update_changelog) sends nothing new.
After publish_changelog, optionally add_comment on each linked post with the changelog URL.
Update or write the help article for the feature (below).
5. Keep the help center current
list_help shows collections and every article with its status. search_help_articles finds published ones by keywords.
Before writing, search. Update what exists (update_help_article) rather than adding a near-copy.
New article: create_help_article. Phrase the title as the question people ask, give a one-line excerpt, write a markdown body with short steps, and set collection_id. It stays a draft until publish_help_article.
Group articles with create_help_collection (icons: book, rocket, card, plug, shield, gear, users, chat, lightbulb, code).
Repeated questions in feedback (the same how-to asked in posts) are a cue for a new article. Reply on those posts with its URL.
Tool map
reference/tools.md lists every tool by area. The server's own prompts (weekly-triage, setup-openheard, ship-and-announce, reply-to-feedback) walk through the same flows.
Openheard next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
Capture conversations and decisions into structured Notion pages; use when turning chats/notes into wiki entries, how-tos, decisions, or FAQs with proper linking.
Cross-reference recently shipped tickets against your help center and changelog to detect outdated articles, missing documentation, and undocumented features.
Builds categorized release notes for a Plane release pull request from its commits and writes them into the PR description, for both the plane-cloud and plane-ee repos.
A skill your agent uses when code-index-mcp implementation is complete and a version bump, release notes, tag, package publication, or GitHub release is being prepared.
Run an openheard feedback board through its MCP server - connect it, set up a workspace from a website, install the widget in this codebase, triage and dedupe feedback, reply, ship and announce…. Openheard is an agent skill from Heilonng23/openheard. Run an openheard feedback board through its MCP server - connect it, set up a workspace from a website, install the widget in this codebase, triage and dedupe feedback, reply, ship and announce changelog entries, keep the help center current.
When should I use Openheard?
Openheard fits situations like: the user mentions openheard; feedback triage; feedback widget.
How do I install Openheard in Claude Code?
Run `npx skills add Heilonng23/openheard --skill openheard -a claude-code`. Or copy the skill folder (skills/openheard in Heilonng23/openheard) into .claude/skills/openheard in your project. Claude Code loads it when a task matches its description.
How do I install Openheard in Codex?
Run `npx skills add Heilonng23/openheard --skill openheard -a codex`. Or copy the skill folder (skills/openheard in Heilonng23/openheard) into .agents/skills/openheard in your project. Codex loads it when a task matches its description.
Can I use Openheard in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Heilonng23/openheard --skill openheard -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/openheard, .gemini/skills/openheard, .github/skills/openheard and .opencode/skills/openheard in your project.
What does Openheard need to run?
Going by SKILL.md and its folder, Openheard needs the command-line tools its instructions call (claude) and credentials named OPENHEARD_KEY. Our summary lists: A credential in OPENHEARD_KEY.
Does Openheard access the network?
SKILL.md names 1 domain. In commands or code: openheard.com; the agent is likely to contact it when it follows the instructions. This is read from the text; nothing was executed.
Is Openheard safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Openheard use?
Openheard is published under the AGPL-3.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Openheard use?
About 1.5k tokens (SKILL.md is roughly 6.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Openheard?
Skills that share tags, products or a category with Openheard: Notion Knowledge Capture (JetBrains/skills, 364 stars), Update Knowledge Base (amplitude/builder-skills, 159 stars), Release (PrefectHQ/fastmcp, 28k stars) and Plane Release Notes Generator (makeplane/plane, 61k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Openheard?
Heilonng23 (a GitHub user) maintains it in Heilonng23/openheard, which has 175 GitHub stars. The repository was last updated on October 3, 2026.
Source: Heilonng23/openheard on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.
