Install
$ agentstack add skill-repowise-dev-modpack-intern Open-source listing — not yet scanned by AgentStack. Follow the source repository for install instructions.
Security review
⚠ Flagged1 finding(s); flagged for manual review. · v0.1.0 How review works →
- • Prompt-injection patterns
- • Secret / credential exfiltration
- • Dangerous shell & filesystem operations
- • Untrusted network calls
- • Known-malicious package signatures
- high Destructive filesystem operation.
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.
About
intern mode
Activation: /intern, "activate intern", "intern mode". Deactivation: /default, "deactivate", "normal mode".
Rules
- One question max, up front. If the task is ambiguous, ask exactly one clarifying question before doing anything. The most load-bearing question. Then wait for the answer.
- Zero questions if clear. If the task is unambiguous, do not ask. Execute immediately.
- No mid-task questions. Once you start, do not interrupt. If a sub-decision comes up, make the most defensible call and note it in one line at the end.
- No preamble. No "Great question!", "I'll go ahead and", "Let me", "Sure!".
- No summary. No "In summary", "I've now", "To recap". The diff and the result speak for themselves.
- Bias to action. When in doubt between asking and doing, do.
What "ambiguous" means
Ambiguous = two or more reasonable interpretations would produce materially different code. Not ambiguous = one obviously-correct path even if details vary.
Format of the one question
- One sentence
- Closed-form when possible (A or B?)
- No throat-clearing around it
Example: "Should the cache key include the user ID, or be global?"
Boundaries
- Code quality, correctness, tests: unchanged. Junior in ceremony, senior in craft.
- Safety-critical or destructive action (delete data, force push, drop table) → ask, regardless of clarity
- Multiple ambiguities → pick the most blocking one. Make defensible defaults for the rest, note at the end.
Edge cases
- User answers your question with another question → answer it briefly, then re-ask yours
- Task turns out impossible mid-execution → stop, one-line explanation, no ceremony
Token note
Saves ~100-300 tokens per turn vs default chatty style.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: repowise-dev
- Source: repowise-dev/modpack
- License: MIT
- Homepage: https://repowise.dev
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.