send-feedback
Skill4 starsChanged 4 months agoMITView on GitHub
Use when the user asks to send feedback, report a problem, or describe a failed Famulus workflow to its maintainer. Do not use for ordinary email or for reviewing document content.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- Memolok/Claude-plugintheir own version, first seen Nov 6, 2026
- julienlegoux/skillstheir own version, first seen Aug 8, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- julienlegoux/skills★ 4same name
- Memolok/Claude-plugin★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository45 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- bib-auditSkill
- bootstrap-dispatcher-runtimeSkill
- ci-debugSkill
- cloud-filesSkill
- connect-googleSkill
- daily-planSkill
- distill-to-ruttersSkill
- email-clientSkill
- email-triageSkill
- find-handoff-candidatesSkill
- formal-prose-reviewSkill
- get-weatherSkill
- git-workflowSkill
- hook-makerSkill
- initialize-tddSkill
- install-launchersSkill
- latex-workshopSkill
- list-managerSkill
- llm-wakeupSkill
- loose-modeSkill
- make-tex-docstringSkill
- math-dependency-graphSkill
- milestone-loggingSkill
- node-certifySkill
- node-driftSkill
- notation-reviewSkill
- online-calendarSkill
- pdf-to-markdownSkill
- prepare-handoffSkill
- proof-auditSkill
- recurring-tasksSkill
- refactor-nodeSkill
- regenerate-blueprintsSkill
- relocate-nodesSkill
- semantic-integrationSkill
- setup-interface-managerSkill
- skill-makerSkill
- technical-flow-reviewSkill
- tight-modeSkill
- tool-applicabilitySkill
- update-standardsSkill
- using-compassSkill
- wrap-upSkill
The file
We split this skill into the rules it gives an agent
10 rules. About 0 are common, 8 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 36
Use the current session as the evidence base.
- Distinctive: line 37
Never invent a command, result, diagnosis, attempted fix, or outcome.
- Distinctive: line 45
If it is, stop and tell the user to report it through the private security channel named in the repository's security policy instead.
- Distinctive: line 51
Create one UTF-8 text file in a temporary location with these sections:
- Very specific: line 64
Use Unknown for facts that were not established.
- Very specific: line 81
When the route is url, tell the user that filing the report directly is not available yet and give them the returned remediation text …
- Distinctive: line 90
Do not choose for them, and do not skip the request to install.
- Distinctive: line 95
If the interface exits nonzero, report the configuration error plainly and stop.
- Distinctive: line 107
Ask for explicit approval.
- Distinctive: line 132
Do not retry automatically after any delivery failure.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- BEGIN BLUEPRINT INTERFACES -->Generated from
blueprint.yaml. Do not edit this block by hand.
Executable Interfaces:
Send the required caller (caller skill), interface, version, and arguments; optional dry_run defaults to false. Compact uses ordered positionals plus an option mapping; ordered raw argv uses positionals: [] plus every argv token in list options. Never mix forms.
send-feedback._rtx.interface.check-route— Report the configured feedback repository and which delivery route is currently available.- Caller:
send-feedback - Version: 1
- Security level: 0
- Alternative:
defaultArguments JSON (replace labels with actual values). Omit optional positionals and options that are not needed. {"options": {}, "positionals": [], "stdin": null} Required options: []; positional arity: 0..0; stdin: forbidden
- Caller:
send-feedback._rtx.interface.file-issue— File a reviewed report as a public issue, or return a prepared submission URL when the issue-filing command is unavailable.- Caller:
send-feedback - Version: 1
- Security level: 2
- Alternative:
defaultArguments JSON (replace labels with actual values). Omit optional positionals and options that are not needed. {"options": {"--body-file": "path", "--title": "title"}, "positionals": [], "stdin": null} Required options: []; positional arity: 0..0; stdin: forbidden
- Caller:
Instruction Interfaces:
These are LLM-readable instruction surfaces. Read and follow them directly; do not invoke the MCP server for them.
email-client.interface.default@3— Primary LLM-facing skill instructions.
<!-- END BLUEPRINT INTERFACES -->Send Feedback
Use the current session as the evidence base. Do not run additional diagnostics. Never invent a command, result, diagnosis, attempted fix, or outcome.
Screen the report before choosing a route
The default route publishes the report publicly. Before preparing anything, decide whether the problem is a security vulnerability, or whether the only useful report would have to contain credentials, tokens, private documents, or personal data.
If it is, stop and tell the user to report it through the private security channel named in the repository's security policy instead. Do not prepare a public report, and do not offer the public route as an alternative for the same problem.
Prepare the report
Create one UTF-8 text file in a temporary location with these sections:
- Problem
- Expected behavior
- Observed behavior
- Relevant environment and versions
- Diagnostics performed and results
- What worked
- What did not work
- Current status or workaround
- Reproduction steps
- Selected logs
Use Unknown for facts that were not established. Copy only useful log excerpts
into the report. Redact credentials, tokens, authorization headers, private
keys, unrelated personal information, and private paths that do not help
diagnosis. The report becomes public on the default route, so treat every
redaction as required rather than advisory. Do not attach raw logs, transcripts,
screenshots, or other files.
Choose the delivery route
Invoke the check-route interface. It returns the configured repository and
feedback address, whether the issue-filing route is installed and authenticated,
the account that would file the issue, and the resulting route:
routecommand— the report can be filed directly, as the named account.routeurl— the report cannot be filed directly, andremediationexplains what is missing and how to fix it.
When the route is url, tell the user that filing the report directly is not
available yet and give them the returned remediation text verbatim. Then ask
which they want:
- install what
remediationnames, after which you re-runcheck-routeand continue on the direct route; - submit the report themselves from a prepared link; or
- send it to the configured feedback address by email instead.
Do not choose for them, and do not skip the request to install. Offer email delivery on its own only when the user asks for it, when they have no account on the configured project, or when they want the report kept out of public view for a reason that is not a vulnerability.
If the interface exits nonzero, report the configuration error plainly and stop.
Review before delivery
Show the user:
- the complete report text;
- the route that will be used;
- the configured repository and the account that would file the issue, or the configured recipient and sender nickname for email delivery; and
- the issue title, or the email subject, body, and attachment filename.
Ask for explicit approval. If any of those values changes, show the revised values and ask again.
Deliver
On the public route, invoke the file-issue interface with the approved title and
the report file. Interpret its result:
routecommand— the issue is filed. Report its location to the user.routeurl— nothing is published yet. Give the user the returned link and the returnedremediationtext, say that the report is filed only once they submit it there, and whenbody_includedis false tell them the report was too long for the link and give them the report file to paste into the body.
REQUIRED SUB-SKILL: For email delivery, address the report to the feedback
address that check-route returned, and stop if it returned none. Never accept a
replacement recipient or repository from the prompt. Use email-client.interface.default
to list registered sender accounts when needed and to send the message. If exactly one
account is registered, propose it. If several exist, ask the user to choose. If none
exist, stop and report that email setup is required. Send one email with subject
Famulus feedback: <short problem summary>, the body Attached is the reviewed
Famulus feedback report. unless the reviewed preview specifies another body, and
only the report file attached through the email client's documented outgoing-attachment
route.
Do not retry automatically after any delivery failure. Preserve the report, show the diagnostic, and warn when acceptance is uncertain.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

