Install
$ agentstack add skill-ph3on1x-wisci-select ✓ 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
/select — Context Loader
Load exactly the context the current task needs. Every stored file is validated by staleness detection before it enters the window — outdated content is stripped, never loaded.
Store state
!python3 "${CLAUDE_PLUGIN_ROOT}/scripts/wisci.py" scan
> If the scan above shows an error or an unexpanded variable, run the bundled script manually: it is at [scripts/wisci.py](scripts/wisci.py) relative to this skill's directory. Run python3 scan with Bash.
The scan classifies every file in .wisci/ (fresh / stale / broken), lists what changed, and enumerates handoff streams. Trust it — do not re-derive staleness with your own git commands.
Mode Detection
Check $ARGUMENTS:
- Empty or blank: execute bare mode — read and follow [references/bare-mode.md](references/bare-mode.md)
- Has content: execute targeted mode — read and follow [references/targeted-mode.md](references/targeted-mode.md)
Staleness Action Rules
Apply to every context file before loading:
- Fresh: load as-is.
- Stale: load with stale sections stripped. A section "depends on" a reference if that reference's path appears in the section body.
- Strip sections whose dependent references appear in the scan's
changedlist. - If a changed reference appears only in
## Referencesand in no section body, strip nothing — prepend instead:
`` > **Note:** Referenced file has changed since this context was written. ``
- When sections are stripped, prepend:
`` > **Note:** N sections stripped — their referenced files changed. Consider /write to refresh. ``
- Broken: do not load. Report: " references code that no longer exists. Delete it or refresh with /write."
- Handoff leaves older than 7 days (scan
age_days): warn before loading — "handoff is N days old; its next steps may be outdated." - Files without a
## Referencesmanifest have no trackable references — treat as fresh.
Output Contract
End every invocation with a load report:
Loaded: files (~ tokens). Stripped: stale sections. Skipped: .
Estimate tokens as total loaded bytes / 4. The point: the user sees what this load cost their context window.
Key Constraints
- Runs inline (not forked) — loaded context must land in the main window. Targeted mode may spawn child agents for exploration.
- Never load stale or broken content; the stripping rules are not optional.
- The handoff index (
.wisci/handoff.md) is script-generated — read it for routing, never edit it.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: ph3on1x
- Source: ph3on1x/wisci
- 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.