Skills explore-code
๐Ÿ“ฆ

explore-code

Content revision r2 High Risk โš™๏ธ External commands๐Ÿ“ Filesystem access

Plan Auditable Exploratory Code Changes

Deep learning experiments can blur candidate changes with trusted baseline work. This skill guides isolated adaptations and creates rollback-aware implementation records.

Supports: Claude Codex Code(CC)
โš ๏ธ 38 Poor

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request
Review the Skillstore skill "explore-code" from https://skillstore.io/skills/lllllllama-explore-code.md and its manifest at https://skillstore.io/api/skills/lllllllama-explore-code/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Test it

Using "explore-code". Plan a reversible LoRA insertion for the image encoder on an isolated branch.

Expected outcome:

  • Candidate targets: encoder module, model registry, and experiment configuration.
  • Minimal plan: add adapter wiring, expose rank settings, and preserve the baseline route.
  • Validation: syntax parsing, import resolution, constructor checks, and one short candidate run.
  • Rollback: remove adapter registration and restore the original configuration.

Using "explore-code". Assess a source-anchored segmentation head transplant.

Expected outcome:

The plan maps the head to likely model and configuration files, protects metric code, and records unresolved comparability evidence.

Using "explore-code". Summarize an exploratory backbone adaptation.

Expected outcome:

The summary identifies changed modules, research anchors, planned checks, rollback steps, and why results remain candidate evidence.

Security Audit

High Risk
v7 โ€ข 8/8/2026 Open versioned report

All 25 static alerts are false positives caused by Markdown formatting, ordinary Python syntax, fixed documentation paths, and conservative policy wording. However, a separate high-severity issue exists because scripts/write_outputs.py dynamically executes an unbundled Python module from outside the skill directory.

5
Files scanned
465
Lines analyzed
0
Review items
0
False positives ignored

Confirmed security concerns (1)

High
Unbundled External Python Module Execution
The output wrapper resolves a module three parent directories above the skill and executes it through importlib. That external code is outside this audited skill bundle.
The code directly constructs an out-of-package path and calls exec_module on it. The external module content is not included in the audited files.
Audited by: codex View Audit History โ†’
Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report
Security Assessment

Copy report link

https://skillstore.io/skills/lllllllama-explore-code/audits/7?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/lllllllama-explore-code/security.svg)](https://skillstore.io/skills/lllllllama-explore-code?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/lllllllama-explore-code?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/lllllllama-explore-code/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/lllllllama-explore-code.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>
Academic citations (APA ยท BibTeX ยท CFF)

APA citation

lllllllama. (2026). explore-code security audit report (audit version 7) [Author version unspecified]. Skillstore. https://skillstore.io/skills/lllllllama-explore-code/audits/7

BibTeX citation

@techreport{lllllllama-lllllllama-explore-code-2026, author = {lllllllama}, title = {explore-code security audit report (audit version 7)}, institution = {Skillstore}, year = {2026}, number = {7}, url = {https://skillstore.io/skills/lllllllama-explore-code/audits/7}, note = {Author version unspecified} }

CITATION.cff

cff-version: 1.2.0 message: "If you use this Skill, cite its author and this versioned security audit report." title: "explore-code security audit report (audit version 7)" version: "unspecified" type: report authors: - name: "lllllllama" date-released: "2026-08-08" url: "https://skillstore.io/skills/lllllllama-explore-code/audits/7" identifiers: - type: other value: "skillstore:lllllllama-explore-code:audit:7" description: "Skillstore immutable audit report identifier"

Skillstore Score

Why this score Evidence Confidence: Medium
64
Architecture
85
Maintainability
87
Content
67
Community
83
Spec Compliance

What You Can Build

Adapt a Research Backbone

Plan a minimal backbone adaptation while preserving configuration wiring and rollback steps.

Review a Module Transplant

Map a source-anchored module into likely code and configuration targets before implementation.

Prepare an Adapter Experiment

Structure a reversible LoRA or adapter change with smoke checks and candidate-level documentation.

Try These Prompts

Find Candidate Targets
On my isolated experiment branch, identify likely files for this authorized adaptation: [change]. Keep the trusted baseline unchanged.
Plan a Reversible Adaptation
Create a minimal patch plan for [module or backbone change]. Include target files, configuration wiring, rollback steps, and planned smoke checks.
Plan from Research Inputs
Use [variant specification], [idea card], and [analysis hints] to plan an isolated adaptation for [task family]. Preserve [research anchor].
Audit a Candidate Implementation
Review this exploratory implementation against [research anchor]. Summarize changed surfaces, comparability risks, rollback actions, and evidence still needed before trusted claims.

Best Practices

  • Use an isolated branch or worktree and record the trusted research anchor.
  • Provide structured variant, idea, and analysis inputs when available.
  • Keep adaptations small, reversible, and clearly labeled as exploratory.

Avoid

  • Do not modify the trusted baseline by default.
  • Do not use candidate results to claim verified reproduction or contribution.
  • Do not use this skill for broad refactors or unsupported implementations from scratch.

Frequently Asked Questions

Does this skill modify the trusted baseline?
No. It requires authorized work on an isolated branch or worktree.
Can it run training experiments?
No. It plans smoke validation and may hand execution to a separate training skill.
Which files can the planner inspect?
It ranks Python, YAML, JSON, TOML, and INI paths while skipping common output and temporary directories.
Does it verify scientific improvements?
No. It treats all exploratory outcomes as candidate evidence until separate reproduction and comparison work is complete.
What inputs improve the plan?
A research anchor, experiment branch, task family, variant specification, idea card, and structural analysis improve targeting.
What records does it expect?
It expects a changeset, scientific changelog, comparability report, top-runs summary, and status record under explore_outputs.

Developer Details

Author

lllllllama

License

MIT

Skillstore revision

r2

Version notice

The author did not declare a version.

Ref

656be3040aef5c047555a908cd5c695d22a4a548

Maintenance freshness

8/8/2026

Usage

5 downloads ยท 0 views

File structure

๐Ÿ“ agents/

๐Ÿ“„ openai.yaml

๐Ÿ“ references/

๐Ÿ“„ explore-policy.md

๐Ÿ“ scripts/

๐Ÿ“„ plan_code_changes.py

๐Ÿ“„ write_outputs.py

๐Ÿ“„ SKILL.md