IR PDF Downloader

Discover and download static PDF files from Investor Relations (IR) sites, including annual reports and quarterly result PDFs. Use when the task is to find a...

Security Scan

The skill's code and instructions are consistent with its stated purpose (discovering and downloading static IR PDFs); it asks for no credentials, has no install script, and only performs web queries and local file writes.

Purpose & Capability

Name/description match the implementation: scripts search Wayback, probe IR domains, query SEC EDGAR for PDF links and download static PDF URLs. Included issuer hints (references/issuers.json) are used by discovery logic and align with the described scope.

Instruction Scope

SKILL.md instructs running the included scripts to discover or download PDFs and explicitly limits scope (no structured XBRL/finance). The scripts perform network requests only to web.archive.org, sec.gov, and target IR domains, and write downloaded PDFs to disk. They do not read unrelated system files or request unrelated environment variables.

Install Mechanism

No install spec is provided (instruction-only). The code requires the standard 'requests' Python package and prints guidance to install it if missing — a low-risk dependency. Nothing is downloaded from arbitrary URLs during install.

Credentials

The skill requests no environment variables, secrets, or external credentials. All network access is explained by the stated purpose (Wayback, SEC EDGAR, IR domains). There are no config paths or tokens requested that would be disproportionate.

Persistence & Privilege

always is false and the skill is user-invocable. The scripts write downloaded PDF files to a local 'downloads' directory (or a user-specified output) — this is expected behavior and does not modify other skills or system configuration.

Assessment

This skill appears coherent with its purpose, but before installing/running consider: (1) it will make outbound HTTP requests to archive.org, sec.gov and arbitrary IR domains and will write files to disk — run in a controlled environment if you are cautious; (2) the downloader sets Referer and other headers to access PDFs behind basic protections (intended to retrieve IR PDFs) — do not use it to bypass access controls for non-public content; (3) review the included Python scripts (they are short and readable) before use and ensure you have the 'requests' package installed; (4) verify you have permission to download large numbers of files from target sites to avoid rate-limiting or terms-of-service issues.

Use this skill for documents, not financial metrics.

Scope

• Find likely PDF URLs from an IR domain
• Search Wayback Machine for archived PDF links
• Use SEC EDGAR only as a PDF discovery source
• Download known static PDF URLs with proper headers

Do not use this skill for

• Revenue / net income / EPS analysis
• SEC XBRL/companyfacts queries
• Filing data normalization

Use sec-finance for those.

Core commands

# Download a known static PDF URL
python3 scripts/download_ir_pdf.py "https://ir.jd.com/static-files/..."

# Discover likely PDFs from an IR domain
python3 scripts/find_ir_pdf.py --domain ir.baidu.com --year 2024

# Search by company alias (uses shared issuer map when available)
python3 scripts/find_ir_pdf.py --company Alibaba

# Search only Wayback
python3 scripts/find_ir_pdf.py --domain ir.alibabagroup.com --sources wayback

Workflow

1. If the PDF URL is already known, use download_ir_pdf.py directly.
2. If only the company or IR domain is known, use find_ir_pdf.py to discover likely URLs.
3. If discovery succeeds, pass the resulting URL to download_ir_pdf.py.
4. If the user actually wants structured financial numbers rather than the document itself, stop and switch to sec-finance.

Notes

• Issuer hints for aliases, IR domains, and validated CIKs live in references/issuers.json.
• Keep this skill focused on PDF discovery and download.
• Do not duplicate structured-finance guidance here.