Fähigkeiten last30days
📦
last30days
v3.30.0 Inhaltsrevision r1 Kritisch 📁 Dateisystemzugriff⚙️ Externe Befehle🔑 Umgebungsvariablen🌐 Netzwerkzugriff⚡ Enthält Skripte
Research Recent Community Discussions
Recent discussions are scattered across social platforms and public websites. This skill gathers recent evidence, compares engagement, and produces cited research with source coverage.
Unterstützt: Claude Codex Code(CC)
Diesen Skill nicht automatisch installieren.
Die maßgebliche Richtlinie verlangt vor jeder Installationsaktion eine Prüfung durch einen Bediener.
Mit meinem Agent installieren
Kopieren Sie diese Anfrage in Ihren Agent. Sie enthält die maßgebliche Skill-Seite und das Manifest.
Agent-Anfrage
Review the Skillstore skill "last30days" from https://skillstore.io/skills/mvanhorn-last30days.md and its manifest at https://skillstore.io/api/skills/mvanhorn-last30days/manifest. Verify the artifact. Do not auto-install. Inspect the skill and report your findings, then wait for an operator or manual installation decision.Ihr Agent sollte weiterhin seinen Plan anzeigen und alle von der Sicherheitsrichtlinie verlangten Bestätigungen anfordern.
Agent-lesbare Ressourcen
Verwenden Sie diese Links, wenn ein KI-Agent, Crawler oder Skript sauberen Kontext benötigt, statt die vollständige Seite zu lesen.
Testen
„last30days“ wird verwendet. Research recent feedback about a developer tool.
Erwartetes Ergebnis:
- Illustrative structure only, not live research results.
- Key themes: recurring praise, reported problems, and areas with insufficient evidence.
- Evidence: attributed comments, source links, dates, and engagement where available.
- Coverage: sources that completed, returned partial results, or could not be reached.
„last30days“ wird verwendet. Compare two products and prepare a local brief.
Erwartetes Ergebnis:
- Illustrative structure only, not a completed comparison.
- Quick verdict: evidence-backed tradeoffs for the requested use case.
- Per-product findings: strengths and complaints supported by separate citations.
- Deliverable: a local HTML brief, with a saved location only when creation succeeds.
Sicherheitsaudit
Kritischv1 • 25.10.2026 Versionsbericht öffnen
All 400 presented findings were adjudicated; many are lexical false positives, but remote shell installers and browser credential access remain real risks. Semantic review identified consent bypass, cross-provider token disclosure, and deliberate scanner-visibility reductions; no audit-directed prompt injection evidence was found. The analyzer omitted 4,034 additional matches, which remain unconfirmed and require manual review before automatic publication. Static review was capped at 400/4434 representative findings; omitted static matches are unconfirmed, so automatic publishing stays disabled until manual review.
153
Gescannte Dateien
76,254
Analysierte Zeilen
8
Prüfelemente
0
Falschmeldungen ignoriert
Bestätigte Sicherheitsbedenken (18)
Kritisch
Pipe to shell pattern
- **Grok CLI (no X credential):** install with `curl -fsSL https://x.ai/cli/install.sh | bash`, then
The setup guidance recommends downloading an unpinned remote installer directly into Bash. Following it executes changing network content without integrity verification.
Kritisch
Pipe to shell pattern
"install the Grok CLI: curl -fsSL https://x.ai/cli/install.sh | bash, "
The setup guidance recommends downloading an unpinned remote installer directly into Bash. Following it executes changing network content without integrity verification.
Kritisch
Pipe to shell pattern
Install: curl -fsSL https://x.ai/cli/install.sh | bash (or npm i -g @xai-official/grok)
The setup guidance recommends downloading an unpinned remote installer directly into Bash. Following it executes changing network content without integrity verification.
Kritisch
Pipe to shell pattern
"install the Grok CLI: curl -fsSL https://x.ai/cli/install.sh | bash, then run `grok login`",
The setup guidance recommends downloading an unpinned remote installer directly into Bash. Following it executes changing network content without integrity verification.
Kritisch
Pipe to shell pattern
"reinstall the Grok CLI: curl -fsSL https://x.ai/cli/install.sh | bash; the current binary won't run
The setup guidance recommends downloading an unpinned remote installer directly into Bash. Following it executes changing network content without integrity verification.
Kritisch
Pipe to shell pattern
"install the Grok CLI (curl -fsSL https://x.ai/cli/install.sh | bash) "
The setup guidance recommends downloading an unpinned remote installer directly into Bash. Following it executes changing network content without integrity verification.
Kritisch
Browser credential files
# finder (which also handles the modern Network/Cookies layout).
These paths support real Chromium cookie discovery and decryption across browser profiles. Consent gating reduces exposure, but session credential access remains sensitive and requires explicit authorization.
Kritisch
Browser credential files
same modern ``Default/Network/Cookies`` (Chromium >= 96) and legacy
These paths support real Chromium cookie discovery and decryption across browser profiles. Consent gating reduces exposure, but session credential access remains sensitive and requires explicit authorization.
Alle 18 bestätigten Funde anzeigen
Kritisch
Browser credential files
``Default/Cookies`` probing as the rest of the Chromium family.
These paths support real Chromium cookie discovery and decryption across browser profiles. Consent gating reduces exposure, but session credential access remains sensitive and requires explicit authorization.
Kritisch
Browser credential files
Prefers the modern ``Network/Cookies`` location (Chromium >= 96 moved the
These paths support real Chromium cookie discovery and decryption across browser profiles. Consent gating reduces exposure, but session credential access remains sensitive and requires explicit authorization.
Kritisch
Browser credential files
for rel in ("Network/Cookies", "Cookies"):
These paths support real Chromium cookie discovery and decryption across browser profiles. Consent gating reduces exposure, but session credential access remains sensitive and requires explicit authorization.
Kritisch
Browser credential files
Each location is probed for both the modern ``Network/Cookies`` and legacy
These paths support real Chromium cookie discovery and decryption across browser profiles. Consent gating reduces exposure, but session credential access remains sensitive and requires explicit authorization.
Kritisch
Browser credential files
Parses ~/Library/Cookies/Cookies.binarycookies (unencrypted binary format)
The Safari extractor actually reads browser cookie stores and returns matching session values. Browser selection gates its caller, but this remains privileged access to reusable authentication material.
Kritisch
macOS security command
Calls `security find-generic-password` which may trigger a system dialog
The subprocess retrieves a browser Safe Storage passphrase using security find-generic-password -w. This enables cookie decryption and exposes browser authentication material to the process.
Kritisch
[HEURISTIC] DANGEROUS COMBINATION: Network + Credentials + Evasion techniques
This combination suggests data exfiltration with anti-detection
Credential-backed networking coexists with explicit scanner-avoidance comments in scripts/lib/env.py and installed runtime vendor exclusions in .skillignore. This weakens automated visibility into sensitive behavior.
Hoch
Sidecar Cookie Reads Bypass Browser Refusal
On extras hosts, credential discovery calls agentcookie before checking browser consent. BROWSER_CONSENT=false or FROM_BROWSER=off does not prevent this session-cookie read.
The extras branch calls read_x_cookies without consulting browser refusal. The sidecar checks only AGENTCOOKIE=off and binary availability before retrieving the pair.
Hoch
Legacy Token Fallback Crosses Provider Boundaries
When SCRAPECREATORS_API_KEY is absent, get_tiktok_token returns APIFY_API_TOKEN. SKILL.md explicitly states that this legacy path sends the Apify credential to ScrapeCreators.
The fallback expression selects a different provider's token, and the security disclosure confirms its destination. Disclosure does not preserve provider-specific credential isolation.
Hoch
Credential Scanner Avoidance and Runtime Exclusion
The environment helper explicitly avoids direct credential-key matches in install-time scanning. The ignore file also excludes the installed runtime X client from scanning.
The comments explicitly explain scanner avoidance and identify the vendor directory as installed but scan-excluded. This reduces audit coverage without proving malicious exfiltration intent.
Elemente der Fähigkeitsprüfung (8)
Dies sind echte lokale Fähigkeiten, die für diese Fähigkeit erwartet werden können; daher müssen sie überprüft werden, werden jedoch nicht als bestätigtes bösartiges Verhalten gezählt.
Hoch
Python subprocess.run
result = subprocess.run(
This subprocess retrieves live X session credentials from the sidecar. Activation can occur without browser consent, so the risk is unauthorized credential access rather than shell injection.
Hoch
Python subprocess.run
result = subprocess.run(
The subprocess retrieves a browser Safe Storage passphrase using security find-generic-password -w. This enables cookie decryption and exposes browser authentication material to the process.
Mittel
Python os file operations
os.makedirs(recipe["profile_dir"], exist_ok=True)
The helper reuses a predictable /tmp browser-profile directory without checking ownership or enforcing private permissions. A precreated directory can cross local trust boundaries during a sensitive login.
Mittel
Temp directory access
**Extras-host X login (Linux / Mac mini / Darwin agentcookie sink — a MacBook SKIPS this, and so doe
The setup recipe directs a signed-in Chrome session to a predictable shared-/tmp profile. It does not require a unique private directory or verify preexisting ownership.
Mittel
Temp directory access
- **Extras hosts (Linux / Mac mini / Darwin agentcookie sink) — a MacBook SKIPS this, and so does a
The setup recipe directs a signed-in Chrome session to a predictable shared-/tmp profile. It does not require a unique private directory or verify preexisting ownership.
Mittel
Temp directory access
2. Launch a throwaway login Chrome on the last30days extras port **18800** — this port is the last30
The setup recipe directs a signed-in Chrome session to a predictable shared-/tmp profile. It does not require a unique private directory or verify preexisting ownership.
Mittel
Temp directory access
mkdir -p /tmp/last30days-x-chrome
The setup recipe directs a signed-in Chrome session to a predictable shared-/tmp profile. It does not require a unique private directory or verify preexisting ownership.
Mittel
Temp directory access
DEFAULT_PROFILE_DIR = "/tmp/last30days-x-chrome"
The helper reuses a predictable /tmp browser-profile directory without checking ownership or enforcing private permissions. A precreated directory can cross local trust boundaries during a sensitive login.
Risikofaktoren
📁 Dateisystemzugriff (50)
.skillignore:2 references/comparison.md:71 references/comparison.md:115 references/comparison.md:89 references/comparison.md:59 references/comparison.md:60 references/comparison.md:61 references/comparison.md:66 references/comparison.md:110 references/discovery.md:38 references/discovery.md:42 references/discovery.md:44 references/discovery.md:75 references/followups.md:19 references/followups.md:20 references/followups.md:21 references/followups.md:22 references/followups.md:23 references/followups.md:19 references/followups.md:20 references/followups.md:21 references/followups.md:22 references/followups.md:23 references/grok-bot-x.md:42 references/grok-bot-x.md:42 references/grok-bot-x.md:45 references/research-runbook.md:56 references/research-runbook.md:56 references/research-runbook.md:603 references/research-runbook.md:605 references/research-runbook.md:607 references/runtime.md:28 references/runtime.md:159 references/runtime.md:161 references/runtime.md:12 references/runtime.md:26 references/runtime.md:28 references/runtime.md:159 references/runtime.md:161 references/runtime.md:12 references/runtime.md:43 references/runtime.md:56 references/runtime.md:69 references/runtime.md:75 references/runtime.md:88 references/runtime.md:89 references/runtime.md:92 references/runtime.md:93 references/save-html-brief.md:52 references/save-html-brief.md:202
⚙️ Externe Befehle (50)
references/comparison.md:67-81 references/comparison.md:162-165 references/comparison.md:165-166 references/comparison.md:166-167 references/comparison.md:167-170 references/comparison.md:170-171 references/comparison.md:171-195 references/comparison.md:195 references/comparison.md:207 references/comparison.md:38-41 references/comparison.md:42-45 references/comparison.md:46-49 references/comparison.md:50-53 references/comparison.md:60 references/comparison.md:61 references/comparison.md:93-96 references/comparison.md:100-103 references/comparison.md:104-107 references/comparison.md:110 references/discovery.md:43-50 references/discovery.md:44 references/discovery.md:75 references/discovery.md:43-50 references/grok-bot-x.md:45 references/research-runbook.md:608-612 references/research-runbook.md:587 references/research-runbook.md:605 references/runtime.md:85-121 references/runtime.md:12 references/runtime.md:69 references/runtime.md:72 references/runtime.md:73 references/runtime.md:74 references/runtime.md:75 references/runtime.md:89 references/runtime.md:93 references/runtime.md:111 references/runtime.md:114 references/runtime.md:151 references/runtime.md:85-121 references/runtime.md:131 references/save-html-brief.md:33-99 references/save-html-brief.md:61 references/save-html-brief.md:64-65 references/save-html-brief.md:33-99 scripts/box_chrome_login.py:200 scripts/box_chrome_login.py:2 scripts/box_chrome_login.py:5 scripts/box_chrome_login.py:9 scripts/box_chrome_login.py:14
🔑 Umgebungsvariablen (50)
references/competitors.md:27 references/grok-bot-x.md:11 references/research-runbook.md:48 references/research-runbook.md:50 references/research-runbook.md:52 references/research-runbook.md:54 references/research-runbook.md:60 references/research-runbook.md:120 references/research-runbook.md:174 references/research-runbook.md:322 references/research-runbook.md:527 references/runtime.md:119 references/runtime.md:163 references/setup-wizard.md:62 references/setup-wizard.md:72 references/setup-wizard.md:89 references/setup-wizard.md:90 references/setup-wizard.md:91 references/setup-wizard.md:95 references/setup-wizard.md:137 references/setup-wizard.md:140 references/setup-wizard.md:141 references/setup-wizard.md:142 references/setup-wizard.md:165 references/setup-wizard.md:174 references/setup-wizard.md:195 references/setup-wizard.md:196 references/setup-wizard.md:199 references/setup-wizard.md:214 references/setup-wizard.md:219 references/setup-wizard.md:223 references/setup-wizard.md:232 references/setup-wizard.md:236 references/setup-wizard.md:239 references/setup-wizard.md:241 references/synthesis.md:14 scripts/box_chrome_login.py:197 scripts/evaluate_search_quality.py:198 scripts/evaluate_search_quality.py:200 scripts/evaluate_search_quality.py:202 scripts/evaluate_search_quality.py:312 scripts/evaluate_search_quality.py:313 scripts/evaluate_search_quality.py:314 scripts/evaluate_search_quality.py:315 scripts/evaluate_search_quality.py:320 scripts/evaluate_search_quality.py:199 scripts/evaluate_search_quality.py:201 scripts/evaluate_search_quality.py:203 scripts/evaluate_search_quality.py:320 scripts/evaluate_search_quality.py:49
🌐 Netzwerkzugriff (50)
references/setup-wizard.md:208 references/synthesis.md:188 references/synthesis.md:190 references/synthesis.md:192 references/synthesis.md:195 references/synthesis.md:196 references/synthesis.md:197 scripts/box_chrome_login.py:47 scripts/box_chrome_login.py:119 scripts/box_chrome_login.py:119 scripts/evaluate_search_quality.py:18 scripts/evaluate_search_quality.py:49 scripts/last30days.py:923 scripts/lib/amazon.py:20 scripts/lib/amazon.py:44 scripts/lib/backends.py:305 scripts/lib/bird_x.py:426 scripts/lib/bird_x.py:863 scripts/lib/bird_x.py:72 scripts/lib/bird_x.py:73 scripts/lib/bluesky.py:123 scripts/lib/bluesky.py:171 scripts/lib/bluesky.py:310 scripts/lib/bluesky.py:123 scripts/lib/bluesky.py:171 scripts/lib/bluesky.py:310 scripts/lib/bluesky.py:24 scripts/lib/bluesky.py:25 scripts/lib/bluesky.py:40 scripts/lib/bluesky.py:64 scripts/lib/bluesky.py:290 scripts/lib/bluesky.py:387 scripts/lib/bounded_get.py:130 scripts/lib/bounded_get.py:15 scripts/lib/bounded_get.py:64 scripts/lib/bounded_get.py:121 scripts/lib/chrome_cdp.py:44 scripts/lib/chrome_cdp.py:110 scripts/lib/chrome_cdp.py:77 scripts/lib/chrome_cdp.py:81 scripts/lib/chrome_cdp.py:101 scripts/lib/chrome_cdp.py:103 scripts/lib/chrome_cdp.py:428 scripts/lib/chrome_cdp.py:101 scripts/lib/chrome_cdp.py:103 scripts/lib/devto.py:131 scripts/lib/devto.py:131 scripts/lib/devto.py:39 scripts/lib/devto.py:40 scripts/lib/devto.py:41
Erkannte Muster
Pipe to shell pattern×6Browser credential files×7macOS security command[HEURISTIC] DANGEROUS COMBINATION: Network + Credentials + Evasion techniques
Geprüft von: codex
Diesen Bericht teilen & zitieren
Teile den versionierten Bewertungsbericht, das neutrale Badge, die Einbettungskarte und Zitate. Skillstore berichtet Nachweise, ohne zu entscheiden, ob dieser Skill sicher ist.
Berichtslink kopieren
https://skillstore.io/skills/mvanhorn-last30days/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown-Badge
[](https://skillstore.io/skills/mvanhorn-last30days?utm_source=security_passport_badge)HTML-Badge
<a href="https://skillstore.io/skills/mvanhorn-last30days?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/mvanhorn-last30days/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Einbettungskarte
<iframe src="https://skillstore.io/embed/skills/mvanhorn-last30days.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Wissenschaftliche Zitate (APA · BibTeX · CFF)
APA-Zitat
mvanhorn. (2026). last30days security audit report (audit version 1) [Author version 3.30.0]. Skillstore. https://skillstore.io/skills/mvanhorn-last30days/audits/1BibTeX-Zitat
@techreport{mvanhorn-mvanhorn-last30days-2026,
author = {mvanhorn},
title = {last30days security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/mvanhorn-last30days/audits/1},
note = {Author version 3.30.0}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "last30days security audit report (audit version 1)"
version: "3.30.0"
type: report
authors:
- name: "mvanhorn"
date-released: "2026-10-25"
url: "https://skillstore.io/skills/mvanhorn-last30days/audits/1"
identifiers:
- type: other
value: "skillstore:mvanhorn-last30days:audit:1"
description: "Skillstore immutable audit report identifier"
Varianten vergleichen
2 installierbare VariantenJeder Autor bleibt ein eigener installierbarer Skill. Die empfohlene Variante wird nach Skillstore-Nachweisen bewertet.
Warum diese Variante zuerst steht
Höhere Evidenzsicherheit
Autor / Quelle Skillstore-Score Evidenzvertrauen Skillstore-Nutzung Aktualisiert
sickn33 Empfohlen
sickn33-last30days
Skillstore-Score 38
Evidenzvertrauen Mittel
Skillstore-Nutzung 57
Aktualisiert
2026-09-09
mvanhorn Aktuell
mvanhorn-last30days
Skillstore-Score 38
Evidenzvertrauen Niedrig
Skillstore-Nutzung 3
Aktualisiert
2026-10-25
Skillstore-Score
Warum dieser Score Evidenzvertrauen: Niedrig 82
Architektur
100
Wartbarkeit
87
Inhalt
65
Gemeinschaft
91
Spezifikationskonformität
Was Sie erstellen können
Evaluate Developer Tool Feedback
Compare recent discussions about two frameworks and identify recurring strengths, complaints, and unresolved questions.
Track Product and Competitor Reactions
Collect recent community reactions to a product launch and compare them with named competitors.
Find Evidence for Editorial Topics
Discover current discussion topics and review attributed comments before selecting an article or podcast angle.
Diese Prompts ausprobieren
Explore a Topic
/last30days Research recent discussions about [topic] from the last 30 days. Cite the evidence and state coverage gaps.
Focus on Community Feedback
/last30days Research [product] on Reddit and Hacker News. Group recurring complaints and praise, include attributed comments, and distinguish evidence from interpretation.
Compare Alternatives
/last30days Compare [product A] versus [product B] over the last 30 days. Separate each product's evidence and explain tradeoffs with citations.
Combine Private Notes and Public Evidence
/last30days Research [topic] with my local notes at [directory]. Keep document contents local and produce a local HTML brief without publishing.
Bewährte Praktiken
- Name the topic, date window, and required sources. Inspect coverage before interpreting missing results.
- Use dedicated provider credentials and avoid browser-backed access until the reported consent issues are resolved.
- Verify important claims against cited sources and approve credit spending or public publishing separately.
Vermeiden
- Treating partial retrieval or authentication failure as proof that nobody discussed a topic.
- Pasting unrelated provider credentials or running unverified remote installation commands.
- Publishing private notes or saved research without explicit authorization and an access review.
Häufig gestellte Fragen
Does every source need an API key?
No. Several public sources work without keys. Other sources need credentials, an installed CLI, or explicit authorization.
Does it guarantee complete 30-day coverage?
No. Source access, indexing, rate limits, and account tiers affect the available evidence and date range.
Can research incur charges?
Yes. Optional providers and automatic paid backfills can consume credits. Review configured sources and retry behavior before running.
Can it read browser sessions?
Yes, through optional cookie adapters. This audit found a sidecar consent bypass, so browser refusal alone does not disable every cookie source.
Can I include private local documents?
Yes. The corpus feature processes selected local directories and restricts remote processing and export of document evidence.
Are reports published automatically?
Local saving is supported by default configuration. Hosted sharing requires explicit authorization and may create public pages unless protection is selected.
Entwicklerdetails
Autor
mvanhornLizenz
MIT
Autorenversion
v3.30.0
Skillstore-Revision
r1
Ref.
f235fd0d05aad155ada09449398e238597b5b345
Aktualität der Wartung
25.10.2026
Nutzung
0 Downloads · 0 Aufrufe