How do I install the File Reading skill?

Run npx skillmds@latest add yafet-dev/file-reading in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is File Reading safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with File Reading?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is File Reading free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published File Reading?

yafet-dev (@yafet-dev) published this skill. Their other Agent Skills are listed on their SkillMD profile.

File Reading by yafet-dev — Agent Skill

Name: File Reading
Author: yafet-dev

File Reading

Advanced file read and path traversal testing methodology for bug bounty and application security work. Use when testing or reviewing local file inclusion, arbitrary file read, path traversal, SSRF-to-file-read, PDF/HTML export file inclusion, archive extraction traversal, client download path traversal, container or host file disclosure, metadata file access, and workflows where user-controlled paths, URLs, filenames, or rendering content can expose restricted files.

yafet-dev Updated Sep 21, 2026

File contents

File Reading Testing

Core Posture

Treat file-read bugs as boundary failures between user input and filesystem, URL fetchers, renderers, archives, containers, or local clients. Identify which file namespace is reachable and whose secrets live there.

Priority Patterns

Server-side file reads: LFI, path traversal, template includes, PDF/HTML export inclusion, and image/SVG metadata processing.
SSRF-to-file-read: file://, cloud metadata, internal URLs, localhost services, and renderer fetches.
Archive and upload paths: zip slip, tar traversal, filename normalization, symlink handling, and extraction overwrite.
Client-side file exposure: desktop/mobile path traversal, download attachment writes, local token/config leaks.
Container/host boundaries: worker escape, mounted secrets, /proc, environment files, service configs, and source code.

Assessment Loop

Find path-like inputs: filename, URL, attachment, import, export, template, archive entry, image metadata, PDF HTML, and callback fields.
Determine parser and namespace: server FS, container FS, client FS, cloud metadata, renderer sandbox, or storage bucket.
Test normalization differences: encoded traversal, mixed separators, absolute paths, symlinks, archive paths, URL schemes, and application-decoded transport formats when the client or endpoint indicates they are supported.
Confirm with low-risk known files or controlled markers before sensitive targets.
Assess impact by reachable secrets, credentials, source code, configs, tokens, and cross-tenant files.

High-Value Cues

Family	Ask
Export/render	Can HTML/PDF/SVG cause backend file fetch?
Upload/archive	Can filenames escape intended directories?
URL fetch	Are `file://`, localhost, metadata, or redirects followed?
Client app	Can downloaded files write or read outside the safe folder?
Container	Are host mounts, secrets, or source files readable?

Variant Playbook

Try ../, encoded traversal, double encoding, mixed slash/backslash, absolute paths, null-like suffixes, and path prefix tricks.
Where the application expects an encoded value, compare the exact same benign path in raw, Base64, Base64url, or other observed transport forms. Track decoding before path normalization; a 403 to 200 change alone does not prove a file read.
Test symlinks, zip/tar entries, nested archives, long names, Unicode normalization, and Windows drive paths.
Test renderer inputs: HTML iframe/img/link, SVG external entities, CSS url(), PDF conversion, and metadata.
Compare web, mobile, desktop, worker, and export services.
Follow redirect chains from allowed URL schemes or hosts.

Confirmation Discipline

Strong evidence shows file contents or a controlled marker from outside the intended directory/tenant/sandbox. Rule out path echo, not-found differences, and public files.

References

Read references/advanced-methodology.md only when the task needs deeper traversal, application-decoded transport and canonicalization checks, archive, renderer, SSRF-to-file, client-app, container, confirmation, or remediation checks.

---name: file-readingdescription: Advanced file read and path traversal testing methodology for bug bounty and application security work. Use when testing or reviewing local file inclusion, arbitrary file read, path traversal, SSRF-to-file-read, PDF/HTML export file inclusion, archive extraction traversal, client download path traversal, container or host file disclosure, metadata file access, and workflows where user-controlled paths, URLs, filenames, or rendering content can expose restricted files.---# File Reading Testing## Core PostureTreat file-read bugs as boundary failures between user input and filesystem, URL fetchers, renderers, archives, containers, or local clients. Identify which file namespace is reachable and whose secrets live there.## Priority Patterns- Server-side file reads: LFI, path traversal, template includes, PDF/HTML export inclusion, and image/SVG metadata processing.- SSRF-to-file-read: `file://`, cloud metadata, internal URLs, localhost services, and renderer fetches.- Archive and upload paths: zip slip, tar traversal, filename normalization, symlink handling, and extraction overwrite.- Client-side file exposure: desktop/mobile path traversal, download attachment writes, local token/config leaks.- Container/host boundaries: worker escape, mounted secrets, `/proc`, environment files, service configs, and source code.## Assessment Loop1. Find path-like inputs: filename, URL, attachment, import, export, template, archive entry, image metadata, PDF HTML, and callback fields.2. Determine parser and namespace: server FS, container FS, client FS, cloud metadata, renderer sandbox, or storage bucket.3. Test normalization differences: encoded traversal, mixed separators, absolute paths, symlinks, archive paths, URL schemes, and application-decoded transport formats when the client or endpoint indicates they are supported.4. Confirm with low-risk known files or controlled markers before sensitive targets.5. Assess impact by reachable secrets, credentials, source code, configs, tokens, and cross-tenant files.## High-Value Cues| Family | Ask || --- | --- || Export/render | Can HTML/PDF/SVG cause backend file fetch? || Upload/archive | Can filenames escape intended directories? || URL fetch | Are `file://`, localhost, metadata, or redirects followed? || Client app | Can downloaded files write or read outside the safe folder? || Container | Are host mounts, secrets, or source files readable? |## Variant Playbook- Try `../`, encoded traversal, double encoding, mixed slash/backslash, absolute paths, null-like suffixes, and path prefix tricks.- Where the application expects an encoded value, compare the exact same benign path in raw, Base64, Base64url, or other observed transport forms. Track decoding before path normalization; a `403` to `200` change alone does not prove a file read.- Test symlinks, zip/tar entries, nested archives, long names, Unicode normalization, and Windows drive paths.- Test renderer inputs: HTML `iframe/img/link`, SVG external entities, CSS `url()`, PDF conversion, and metadata.- Compare web, mobile, desktop, worker, and export services.- Follow redirect chains from allowed URL schemes or hosts.## Confirmation DisciplineStrong evidence shows file contents or a controlled marker from outside the intended directory/tenant/sandbox. Rule out path echo, not-found differences, and public files.## ReferencesRead `references/advanced-methodology.md` only when the task needs deeper traversal, application-decoded transport and canonicalization checks, archive, renderer, SSRF-to-file, client-app, container, confirmation, or remediation checks.

Frequently asked questions