Install
$ agentstack add skill-tomzx-agents-create-message ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Write Message
Takes a draft message and improves it by removing negative tone, reformulating with positive language, and appending actionable suggestions where appropriate.
Prerequisites
- Draft message: The message the user wants to send (provided inline or in a file)
- Recipient (optional): Who the message is for (e.g., "a colleague", "a manager", "a client"), used to calibrate tone
- Context (optional): Any background that helps understand the intent (e.g., "they missed a deadline", "I want to decline a meeting")
Steps
- Read the draft message carefully and identify the core intent.
- Scan for negative tone markers: passive-aggression, blame, criticism without constructive framing, defensiveness, dismissiveness, or vague complaints.
- Reformulate each negative passage using positive language that preserves the original intent.
- Identify any places where actionable suggestions would strengthen the message, and add them.
- Review the revised message for tone consistency, clarity, and brevity.
Tone Rewrite Rules
Remove:
- Blame language ("you failed to", "you didn't", "you never")
- Passive-aggressive framing ("as I'm sure you're aware", "per my last email")
- Vague complaints ("this isn't good enough", "this needs to be better")
- Defensiveness ("I was just trying to", "that's not what I meant")
- Dismissive language ("whatever you think", "it doesn't matter to me")
- Sarcasm or backhanded compliments
Replace with:
- Ownership language ("I'd like us to", "we could", "I suggest")
- Specific, observable descriptions ("the report arrived on Friday, past the Wednesday deadline" instead of "you were late again")
- Constructive framing ("here's what would help", "moving forward, I'd recommend")
- Direct requests ("could you", "would it be possible to")
- Collaborative tone ("let's", "how about we")
Actionable Suggestions
Add concrete suggestions when the message:
- Requests a change in behavior (suggest what specifically to do instead)
- Reports a problem (suggest one or more solutions)
- Sets expectations (suggest clear next steps with timelines)
- Expresses disagreement (suggest an alternative approach)
Suggestions should be:
- Specific and concrete (not "do better" but "send the summary by Thursday noon")
- Realistic and achievable
- Framed as options where appropriate ("we could A or B")
- Limited to 1-3 per message to avoid overwhelming the recipient
Output Format
## Improved Message
[The rewritten message in full]
---
## Changes Made
- **Tone adjustments**: [list of negative elements removed and what they were replaced with]
- **Actionable suggestions added**: [list of suggestions added, or "None needed" if the message was already actionable]
- **Preserved intent**: [one sentence confirming the core message remains the same]
Example Usage
Scenario 1: Frustrated follow-up Draft: "You still haven't sent me the report. I've asked three times now and I'm getting tired of chasing you." Recipient: colleague Improved: "I haven't received the report yet. Could you send it by end of day Wednesday? If something is blocking you, let me know and I can help unblock it."
Scenario 2: Declining a request Draft: "I don't have time for this. You should have asked me earlier." Recipient: teammate Improved: "I won't be able to take this on this week due to my current commitments. For future requests, reaching out by Monday would give me the best chance to fit it in."
Scenario 3: Feedback on work Draft: "This design is confusing and doesn't make sense. I can't believe we're considering this." Recipient: design team Improved: "I have some concerns about the design's clarity. Specifically, the navigation flow between steps 2 and 3 feels unclear. Could we explore adding a progress indicator or transitional labels? I'd be happy to walk through a specific example in our next review."
Useful Commands Reference
No CLI commands required. This skill operates on content provided in context.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: tomzx
- Source: tomzx/agents
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.