Install
$ agentstack add skill-onlyterp-hermes-optimization-guide-daily-inbox-triage ✓ 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
daily-inbox-triage — Morning Sweep
Produce a one-screen triage report: what's urgent, what's a decision, what's noise, with a draft reply per actionable item.
Procedure
- Collect unread items from each configured channel within
window:. Cap at 200 items; if over, prioritize starred / mentions / VIP-list senders.
- Classify every item into one of:
urgent— time-sensitive, needs action todaydecision— needs a yes/no/pick from meinfo— FYI; noting what they said is enoughnoise— newsletters, generic updates, obvious marketingspam— confident spam (see [spam-trap](../../security/spam-trap/SKILL.md))
- Summarize per item: one line of "who / what / ask". Keep under 80 chars.
- Draft replies for every
urgentanddecisionitem. Keep replies under 4 sentences. Never include URLs the sender supplied without sanitizing.
- Output as a single markdown message:
``` ## Inbox Triage — {date}, last {window}
### Urgent ({n})
- [email] Alice @ Acme — blocker on staging auth → draft: "{reply}" [/approve 1]
- [slack] #incidents — payment API 500s → draft: "{reply}" [/approve 2]
### Decisions ({n})
- [telegram] @pm — approve Q3 roadmap doc? → draft: "{reply}" [/approve 3]
### FYI ({n})
- {brief one-liners}
### Noise ({n})
- {unsubscribable patterns suggested}
```
- Surface:
- Any item matching a VIP sender pattern (from config) gets escalated regardless of classifier.
- Any item mentioning "urgent", "asap", "incident", "outage", "production" escalates to
urgenteven if classifier disagreed.
- Never:
- Send replies automatically. Approval is required for every outbound.
- Follow links from untrusted senders.
- Summarize attachments without explicit user consent (privacy + prompt-injection risk).
Example config snippet
skills:
overrides:
daily-inbox-triage:
vip_senders: [ceo@, "board@", "@lawyer.example.com"]
escalate_keywords: [urgent, asap, incident, outage, production]
cron:
- name: morning-inbox
schedule: "0 8 * * 1-5"
task: "/inbox-triage 24h"
notify: telegram_dm
Tips
- Keep this skill reading-only by default — it reports, you approve, you reply.
- Pair with [telegram-triage](../telegram-triage/SKILL.md) for same-shape logic on Telegram-only flows.
- Route to cheap models (Flash / Cerebras). You'll run this daily; every penny counts.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: OnlyTerp
- Source: OnlyTerp/hermes-optimization-guide
- 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.