# TrailMeta — Complete Product Reference > Extract your Salesforce org metadata and produce AI-ready markdown for Claude, ChatGPT, Gemini, Cursor, and any AI platform. 7-tier extraction pipeline, 5 export formats, 18+ supported platforms. From $15. TrailMeta connects to your Salesforce org via standard APIs (REST, Tooling, Metadata, Describe), extracts all metadata across 7 tiers (objects, fields, flows, Apex, triggers, LWC, profiles, permission sets, connected apps, named credentials, packages), and produces structured markdown documentation optimized for AI consumption. No installation in your org required. Metadata only — no record data accessed. ## Product Overview ### 7-Tier Extraction Pipeline | Tier | What | Salesforce API | |------|------|----------------| | 1 | Objects + Fields (complete data model) | REST Describe + Tooling API | | 2 | Flows, Apex Classes, Triggers, Validation Rules | Tooling API | | 3 | Lightning Web Components | Tooling API | | 4 | Profiles, Permission Sets, Permission Set Groups | Tooling API + Metadata API | | 5 | Connected Apps, Named Credentials | Tooling API | | 6 | Installed Packages, Org Summary | SOQL + Tooling API | | 7 | AI Enrichment (optional) — two-pass analysis | Google Gemini | ### 5 Export Formats | Format | Best For | Description | |--------|----------|-------------| | Single File | Claude chat, Gemini, any AI chat | Everything in one markdown file with TOC | | Consolidated Pack | Claude Projects, ChatGPT Projects, Custom GPTs, NotebookLM Free | Files merged by category (≤20 files) | | Full Pack | NotebookLM Pro, OpenAI File Search, LangChain, LlamaIndex | Individual files for every component | | IDE Project | Cursor, Claude Code, Copilot, Windsurf, Codex | Full Pack + pre-configured IDE files | | API Bundle | Developers building custom integrations | Full Pack + graph.json + metadata.json | ### Pricing - **Extraction:** $15 flat one-time fee. No subscription. All 5 export formats included. - **AI Enrichment (optional):** $15 base + (Gemini API cost × 1.54). Typical totals by org size: - Small orgs (under 50 objects): ~$20 total - Medium orgs (50-200 objects): ~$25-35 total - Large orgs (200-500 objects): ~$40-60 total - Enterprise orgs (500+ objects): ~$60-100+ total - Price is calculated and shown before you confirm. No hidden costs. ### AI Enrichment (Tier 7) **Pass 1 — Component Intelligence:** Per-item analysis of objects, flows, Apex classes, triggers, profiles. Adds complexity scoring, risk flags, dependency notes, and architecture observations. **Pass 2 — Cross-Cutting Analysis:** Generates new documents from the full org context: - Business Process Maps (end-to-end workflows) - Domain Digests (logical groupings of related components) - Impact Analysis (blast radius for changes) - Execution Order analysis (what fires when) - Org Health Report (technical debt, risks, recommendations) - START HERE guide (entry point for AI consumers) ### Security - Metadata only — no record data, no PII, no customer information ever accessed - AES-256-GCM encryption at rest with per-job encryption keys - TLS 1.3 in transit - Automatic permanent deletion after 48 hours - No Salesforce credentials stored — OAuth tokens used only during active session - No installation in your Salesforce org required - Read-only API access — your org is never modified ### Supported Salesforce Editions Enterprise, Unlimited, Performance, and Developer editions. Requires System Administrator profile or equivalent (API Enabled, View Setup and Configuration, Modify All Data). ## Frequently Asked Questions ### What Salesforce data does TrailMeta extract? TrailMeta extracts metadata only — objects, fields, flows, Apex classes, triggers, LWC components, profiles, permission sets, connected apps, named credentials, and installed packages. No record-level data (contacts, leads, opportunities) ever leaves your org. This covers your complete org architecture across all 7 extraction tiers. ### Which AI platforms does TrailMeta support? TrailMeta produces 5 export formats that cover 18+ AI platforms: Claude, ChatGPT, Gemini, Cursor, Copilot, Windsurf, Codex, NotebookLM, Custom GPTs, Cline, Aider, Microsoft Copilot, LangChain, LlamaIndex, and more. Each format is optimized for the platform's specific constraints — context windows, file limits, and format preferences. ### Is my Salesforce org data secure? Yes. TrailMeta extracts only metadata (schema definitions, not business data). All extracted data is encrypted at rest using AES-256-GCM with per-job encryption keys. Files are automatically deleted after 48 hours. No record data, no PII, no customer information is ever accessed. See our Security Policy for full details. ### What does the AI enrichment tier include? AI enrichment runs two analysis passes over your extracted metadata. Pass 1 analyzes each component individually — complexity scoring, risk flags, and architecture notes. Pass 2 generates cross-cutting insights: org health reports, blast radius impact analysis, business process maps, execution order documentation, domain digests, and a START HERE guide for AI consumers. ### How much does TrailMeta cost? Extraction is a flat $15 one-time fee — no subscription. AI enrichment adds processing cost based on your org size, calculated and shown before you confirm. Typical totals: ~$20 for small orgs (under 50 objects), ~$30 for medium, ~$50 for large, ~$80+ for enterprise. The formula is fully transparent: $15 base + (Gemini API cost x 1.54). ### How is this different from Elements.cloud or Hubbl? Elements.cloud ($495-995/month) and Hubbl produce documentation for human readers — dashboards, wikis, diagrams. TrailMeta produces structured markdown optimized for AI consumption. Different output, different audience, different price point. TrailMeta is $15 one-time vs hundreds per month, and the output is designed for AI tools like Claude, ChatGPT, and Cursor — not for browsing in a browser. ### How is this different from Salesforce MCP connectors? MCP connectors (CData, Windsor, Composio) let AI query your transactional data — records, reports, dashboards. TrailMeta extracts your org architecture — the schema, automation, security model, and code. They answer "what data is in my org?" TrailMeta answers "how is my org built?" You need both for complete AI coverage, but architecture context is the foundation. ### Do I need to install anything in my Salesforce org? No. TrailMeta connects via standard Salesforce APIs using your admin credentials. No managed package, no connected app, no AppExchange installation required. Your org is not modified in any way — TrailMeta only reads metadata through official Salesforce REST, Tooling, and Metadata APIs. ### What about managed package components? Custom and unlocked package components are fully extracted — source code, flow definitions, everything. Managed package objects can be optionally included, but only schemas are available. Apex source, flow definitions, LWC markup, and validation rule formulas in managed packages are encrypted by Salesforce and cannot be extracted by any tool. ### What Salesforce editions and permissions are required? TrailMeta works with any Salesforce edition that supports API access — Enterprise, Unlimited, Developer, and Performance editions. You need a user with System Administrator profile or equivalent permissions including API Enabled, View Setup and Configuration, and Modify All Data (for complete metadata access). Read-only admin profiles work for most extraction tiers. ### How long does extraction take? Extraction typically takes 2-5 minutes for small orgs (under 50 custom objects), 5-15 minutes for medium orgs, and 15-30 minutes for large enterprise orgs. AI enrichment adds another 5-20 minutes depending on the number of components. You can watch progress in real time as each of the 7 tiers completes. ### Can I use TrailMeta output with Cursor, Windsurf, or other AI coding tools? Yes — this is one of the most popular use cases. The IDE Project export format generates your full org context plus pre-configured project files: CLAUDE.md for Claude Code, AGENTS.md for Codex, .cursor/rules for Cursor, .windsurfrules for Windsurf, and .github/copilot-instructions.md for GitHub Copilot. Drop the output into your SFDX project and your AI coding assistant instantly understands your org. ### How is this different from running sfdx force:source:retrieve? SFDX retrieves raw source files — XML metadata, Apex files, flow definitions. An AI cannot easily reason about thousands of XML files. TrailMeta extracts the same underlying data but transforms it into structured, cross-referenced markdown that AI models can actually understand. It also adds relationship mapping, dependency graphs, and optionally AI-generated analysis that raw metadata retrieval cannot provide. ## Help Center ### Getting Started **What is TrailMeta and how does it work?** TrailMeta is a web-based tool that connects to your Salesforce org, extracts all metadata (objects, fields, flows, Apex classes, triggers, LWC, profiles, permission sets, connected apps, named credentials, and installed packages), and produces structured markdown documentation optimized for AI platforms. You connect your org credentials, TrailMeta runs a 7-tier extraction pipeline across Salesforce REST, Tooling, and Metadata APIs, then outputs files you can feed directly to ChatGPT, Claude, Gemini, Cursor, Copilot, or any AI tool. No installation in your org is required. **Which Salesforce editions does TrailMeta support?** TrailMeta works with any Salesforce edition that has API access enabled: Enterprise Edition, Unlimited Edition, Performance Edition, and Developer Edition. Professional Edition does not include API access by default — you would need the API Access add-on. Government Cloud (GovCloud) orgs are also supported as long as API access is available. Sandbox orgs (Developer, Developer Pro, Partial, Full) all work and are a great way to test TrailMeta before running it on production. **What Salesforce permissions do I need to use TrailMeta?** You need a Salesforce user with the following permissions: API Enabled, View Setup and Configuration, and Modify All Data (for complete metadata access across all tiers). A System Administrator profile includes all of these by default. If you cannot use a full admin, a read-only admin profile works for most extraction tiers — you may miss some Metadata API components in Tier 4 (profiles and permission sets) without Modify All Data. TrailMeta never modifies your org; all API calls are read-only. **How long does Salesforce metadata extraction take?** Extraction time depends on your org size. Small orgs (under 50 custom objects, fewer than 20 flows) typically complete in 2-5 minutes. Medium orgs (50-200 objects, 20-100 flows) take 5-15 minutes. Large enterprise orgs (200+ objects, 100+ flows, extensive Apex) may take 15-30 minutes. AI enrichment adds another 5-20 minutes depending on component count. You can watch real-time progress as each of the 7 tiers completes. Extraction is designed to respect Salesforce API rate limits so it will not impact your org performance. **Should I run TrailMeta on a sandbox or production org?** Both work. If you want a complete picture of your production architecture — including all custom objects, flows, Apex, profiles, and integrations — run it on production. If you want to test TrailMeta first or only need the development environment's metadata, use a Full or Partial sandbox. Developer sandboxes work but only contain metadata, not data, which is fine for TrailMeta since it only reads metadata anyway. Many teams run TrailMeta on production first, then use the output to plan sandbox-based development with full org context. ### Extraction & Output **What exactly does TrailMeta extract from my Salesforce org?** TrailMeta runs a 7-tier extraction pipeline. Tier 1: Objects and fields (your complete data model with field types, relationships, picklist values, and formulas). Tier 2: Flows, Apex classes, triggers, and validation rules (all automation and business logic). Tier 3: Lightning Web Components. Tier 4: Profiles, permission sets, and permission set groups (your security model). Tier 5: Connected apps and named credentials (integrations). Tier 6: Installed packages and org summary. Tier 7 (optional): AI enrichment — two-pass analysis that adds org health reports, impact analysis, business process maps, and more. Every component includes cross-references showing how it relates to other components in your org. **Does TrailMeta access my Salesforce records or customer data?** No. TrailMeta extracts metadata only — the structure and configuration of your org, not the data stored in it. It reads object definitions, field schemas, flow configurations, Apex source code, profile settings, and similar structural information. It never queries Account records, Contact records, Opportunity records, or any business data. No PII, no customer information, no transactional data is ever accessed, transmitted, or stored. The Salesforce APIs TrailMeta uses (Describe, Tooling API, Metadata API) return schema information, not record data. **What export formats does TrailMeta offer?** TrailMeta produces 5 export formats optimized for different AI platforms: (1) Single File — everything in one markdown file with a table of contents, ideal for Claude chat, Gemini, or any AI chat interface. (2) Consolidated Pack — files merged by category into 20 or fewer files, designed for Claude Projects, ChatGPT Projects, Custom GPTs, and NotebookLM Free (which have file count limits). (3) Full Pack — individual files for every component plus a README, ideal for NotebookLM Pro, OpenAI File Search, LangChain, and LlamaIndex. (4) IDE Project — Full Pack plus pre-configured project files for Cursor, Claude Code, Copilot, Windsurf, and Codex. (5) API Bundle — Full Pack plus graph.json and metadata.json for developers building custom integrations. **How does TrailMeta handle managed packages from the AppExchange?** Custom and unlocked package components are fully extracted — source code, flow definitions, field schemas, everything. For managed (first-generation) packages installed from the AppExchange, TrailMeta can extract object and field schemas but cannot access encrypted source code. This is a Salesforce platform limitation, not a TrailMeta limitation — Apex source, flow XML, LWC markup, and validation rule formulas in managed packages are encrypted and inaccessible to any external tool. You can optionally include or exclude managed package object schemas from your extraction. **Can I re-run extraction after making changes to my org?** Yes. Each extraction is independent. If you make changes to your Salesforce org — add objects, modify flows, update Apex — you can run TrailMeta again to get an updated snapshot. Each run is a separate $15 extraction. There is no subscription and no ongoing cost. Many teams re-extract quarterly or before major projects (migrations, new feature builds, org audits) to ensure their AI tools have current context. ### AI Platform Guides **How do I use TrailMeta output with ChatGPT?** For ChatGPT conversations: download the Single File format and paste it into your conversation, or attach it as a file. For ChatGPT Projects: download the Consolidated Pack (20 or fewer files) and upload all files to a project — ChatGPT will use them as persistent context across all conversations in that project. For Custom GPTs: upload the Consolidated Pack as knowledge files when configuring your GPT. For OpenAI File Search (Assistants API): use the Full Pack format. Each format is optimized for ChatGPT's specific file and context constraints. See our blog post "How to Feed Your Entire Salesforce Org to ChatGPT" for a step-by-step walkthrough. **How do I use TrailMeta output with Claude?** For Claude conversations: use the Single File format — paste it or attach it as a file. Claude's 200K context window handles even large org exports well. For Claude Projects: upload the Consolidated Pack as project knowledge — Claude will reference it across all conversations in the project. For Claude Code (CLI): use the IDE Project format, which includes a pre-configured CLAUDE.md file that tells Claude Code about your org architecture, naming conventions, and key relationships. Drop the output folder into your SFDX project root and Claude Code immediately understands your org. **How do I use TrailMeta output with Cursor IDE?** Download the IDE Project export format. It includes a .cursor/rules/salesforce-org.mdc file — Cursor's native rules format — pre-populated with your org's objects, fields, flows, Apex classes, naming conventions, and relationships. Place the TrailMeta output in your SFDX project root. When you open the project in Cursor, the AI assistant automatically loads your org context. You can then ask Cursor to write Apex that references your actual objects and fields, debug flows with full awareness of your automation chain, and generate test classes that match your real data model. **How do I use TrailMeta output with Google Gemini or NotebookLM?** For Gemini conversations: use the Single File format — Gemini's large context window (1M+ tokens) handles it well. For NotebookLM Free: upload the Consolidated Pack (20 or fewer files) as sources. NotebookLM will index your entire org architecture and let you ask questions, generate summaries, and even create audio overviews of your Salesforce org. For NotebookLM Pro: use the Full Pack with individual files for more granular source attribution. Gemini and NotebookLM are particularly strong at summarizing complex org relationships and generating executive-level architecture overviews. **How do I use TrailMeta output with GitHub Copilot?** Download the IDE Project export format. It includes a .github/copilot-instructions.md file pre-configured with your Salesforce org context — objects, fields, relationships, automation patterns, and naming conventions. Place this in your SFDX project root. GitHub Copilot reads copilot-instructions.md as workspace-level context, so every code suggestion it makes will be informed by your actual org architecture. This is especially useful for writing Apex triggers, test classes, and LWC components that need to reference your real custom objects and fields. ### AI Enrichment **What does TrailMeta AI enrichment actually do?** AI enrichment runs two analysis passes over your extracted metadata. Pass 1 (Component Intelligence): Each object, flow, Apex class, trigger, and profile is individually analyzed — complexity scoring, risk flags, dependency notes, and architecture observations. Pass 2 (Cross-Cutting Analysis): The enrichment engine generates new documents from the full org context: business process maps that trace end-to-end workflows, domain digests that group related components, impact analysis showing blast radius for changes, execution order documentation showing what fires when, an org health report identifying technical debt and risks, and a START HERE guide that tells AI consumers where to begin. **How is AI enrichment pricing calculated?** AI enrichment cost is calculated transparently before you confirm. The formula: $15 base fee + (Gemini API processing cost x 1.54). The 1.54 multiplier covers infrastructure and margin. Typical totals by org size: small orgs (under 50 objects): ~$20 total. Medium orgs (50-200 objects): ~$25-35 total. Large orgs (200-500 objects): ~$40-60 total. Enterprise orgs (500+ objects): ~$60-100+ total. The exact cost is calculated and displayed based on your specific org's component count before you approve the charge. You always see the price before paying. **Do I need AI enrichment or is extraction alone enough?** Extraction alone ($15) gives you complete, well-structured markdown documentation of your entire org — objects, fields, flows, Apex, triggers, profiles, permissions, integrations, and packages. This is enough for most use cases: feeding context to AI coding assistants, documenting your org for onboarding, or understanding your architecture. AI enrichment adds analysis on top: health reports, risk flags, business process maps, and cross-cutting insights. It is most valuable for org audits, migration planning, technical debt assessments, and when you want AI-generated architectural analysis rather than just raw documentation. ### Security & Privacy **How does TrailMeta protect my Salesforce metadata?** All extracted metadata is encrypted at rest using AES-256-GCM with per-job encryption keys. Data is transmitted over TLS 1.3. Each extraction job is isolated — your data is never mixed with other customers' data. Files are automatically and permanently deleted 48 hours after extraction. TrailMeta does not store your Salesforce credentials; OAuth tokens are used only during the active extraction session. No metadata is used for model training or shared with third parties. See our full Security Policy at /security for complete details. **How are my Salesforce credentials handled?** TrailMeta uses Salesforce OAuth 2.0 for authentication — you authorize TrailMeta through Salesforce's standard login flow. Your username and password are never entered into or stored by TrailMeta. The OAuth access token is used only during the active extraction session and is not persisted after extraction completes. If you prefer, you can also connect using a Salesforce security token with username/password authentication, but OAuth is recommended as the more secure option. **Is TrailMeta compliant with enterprise security requirements?** TrailMeta is designed with enterprise security in mind. Key controls: metadata-only extraction (no record data or PII), AES-256-GCM encryption at rest, TLS 1.3 in transit, per-job isolation, 48-hour automatic data deletion, no credential storage, and read-only API access (your org is never modified). TrailMeta does not require a connected app installed in your org, which simplifies security reviews. For organizations with strict data residency requirements, extracted data is processed and stored in US-based infrastructure. ### Pricing & Billing **How much does TrailMeta cost? Is there a subscription?** TrailMeta uses simple, one-time pricing with no subscription. Extraction: $15 flat fee per extraction — you get complete metadata extraction across all 7 tiers with all 5 export formats included. AI Enrichment (optional): $15 base + AI processing cost based on your org size (typically $5-65 additional, calculated and shown before you confirm). There are no monthly fees, no per-user pricing, no contracts, and no hidden costs. You pay once per extraction. If you need to re-extract after org changes, each run is another $15. **How does TrailMeta compare to Elements.cloud pricing?** Elements.cloud costs $495-995/month depending on your plan, with annual contracts. That is $5,940-11,940 per year. TrailMeta is $15 one-time per extraction. The products serve different purposes: Elements.cloud produces interactive documentation for human readers (dashboards, diagrams, wikis). TrailMeta produces structured markdown optimized for AI consumption. If your goal is to give AI tools (ChatGPT, Claude, Cursor, Copilot) complete context about your Salesforce org, TrailMeta is purpose-built for that at a fraction of the cost. **How does TrailMeta compare to Hubbl Diagnostics?** Hubbl Diagnostics provides org health scoring and visual dashboards for Salesforce admins and architects. It is a human-facing analysis tool. TrailMeta produces AI-ready documentation — structured markdown that AI platforms can consume and reason about. If you need a dashboard to browse your org health in a browser, Hubbl is a good choice. If you need to feed your complete org architecture to Claude, ChatGPT, Cursor, or any AI coding tool so it understands your specific objects, flows, and Apex, TrailMeta is the tool for that. ### Use Cases **How do I use TrailMeta for a Salesforce org audit?** Run TrailMeta with AI enrichment enabled. The enrichment tier generates an Org Health Report that identifies technical debt, unused components, security gaps, overly complex automation, and architecture risks. The Impact Analysis documents show blast radius for potential changes. Business Process Maps trace end-to-end workflows so you can see where automation overlaps or conflicts. Upload the output to Claude or ChatGPT and ask specific audit questions: "What are the riskiest components in this org?" "Which flows have redundant logic?" "What objects have no automation but should?" The AI can reason about your entire org architecture at once. **How do I use TrailMeta for Salesforce migration planning?** Extract your source org with TrailMeta, then feed the output to your AI tool of choice. The AI now understands your complete data model, all automation (flows, triggers, validation rules), security model (profiles, permission sets), and integrations. Ask it to help plan your migration: identify dependencies that must move together, flag components that reference hardcoded IDs, surface automation that may break if object relationships change, and generate a phased migration checklist. For org merges or splits, extract both orgs and feed both extractions to the AI to identify conflicts, overlaps, and compatibility issues. **How do I use TrailMeta for developer onboarding?** New developers joining a Salesforce team face weeks of ramp-up time understanding the org architecture. TrailMeta shortens this dramatically. Extract your org, give the new developer the output files, and have them load it into their preferred AI tool. They can ask natural-language questions: "How does the opportunity-to-invoice flow work?" "What triggers fire on the Account object?" "Which profiles have access to the Revenue_Forecast__c object?" The AI becomes an always-available org expert that can explain any component, trace any dependency, and answer any architecture question instantly. **What is vibe coding for Salesforce and how does TrailMeta help?** Vibe coding means using AI coding assistants (Cursor, Claude Code, Copilot, Windsurf) to write code through natural-language conversation rather than manual coding. The challenge with vibe coding Salesforce: your AI assistant does not know your org. It generates generic Apex referencing standard objects when you have custom objects. It writes triggers without knowing your existing automation chain. TrailMeta solves this by giving your AI coding assistant complete org context. With the IDE Project export, your AI knows your objects, fields, relationships, flows, triggers, and naming conventions — so it writes code that actually works in your org. **How do I map Salesforce technical debt with AI?** Traditional technical debt discovery in Salesforce means clicking through Setup screens one component at a time. TrailMeta extracts everything at once, and the AI enrichment tier specifically identifies: unused custom objects and fields, flows with redundant or overlapping logic, Apex classes not referenced by any trigger or test, overly permissive profiles, validation rules that may conflict, and deprecated API versions in use. Feed the output to Claude or ChatGPT and ask for a prioritized technical debt remediation plan. The AI can see patterns across your entire org that would take a human weeks to discover manually. ### Troubleshooting **TrailMeta cannot connect to my Salesforce org. What should I check?** First, verify your credentials work by logging into Salesforce directly. Then check: (1) Your user has API Enabled permission — go to Setup > Users > your user > Profile and confirm. (2) Your org allows API access — some sandbox configurations restrict API access. (3) If using a custom domain (My Domain), ensure you are entering the correct login URL. (4) If your org uses IP restrictions, your current IP may need to be allowlisted in your Salesforce Login IP Ranges or Network Access settings. (5) Check if your org has active Connected App policies that block new OAuth connections. TrailMeta logs connection errors with specific Salesforce error codes to help diagnose the issue. **My extraction seems to be missing some components. Why?** Common reasons: (1) Permission gaps — if your user does not have Modify All Data, some Metadata API components (particularly detailed profile and permission set data) may be inaccessible. (2) Managed packages — source code in managed packages is encrypted by Salesforce and cannot be extracted. You will see the object schemas but not Apex or flow definitions. (3) Recently created components — if you created components after starting extraction, they will not be included. Re-run extraction to pick them up. (4) Tooling API limits — very large orgs may hit Salesforce query limits during extraction. TrailMeta handles pagination automatically, but extremely large result sets may require multiple runs. **Extraction is taking too long or timing out for my large org. What can I do?** Large enterprise orgs (500+ objects, hundreds of flows, extensive Apex) can take 30+ minutes. If extraction times out: (1) Check your browser has not gone to sleep — keep the tab active during extraction. (2) Try extracting during off-peak hours when Salesforce API response times are faster. (3) If specific tiers are slow, it is usually Tier 2 (flows and Apex — large orgs may have hundreds of each) or Tier 4 (profiles and permission sets — some orgs have 50+ profiles). The extraction engine respects Salesforce rate limits, so slower extraction means your org has a lot of metadata — which is exactly why you need TrailMeta. **I entered my PIN wrong and now the page says "Unauthorized". What happened?** For your protection, we lock the PIN for 15 minutes after 5 failed attempts. This prevents anyone who intercepts your download link from brute-forcing the 6-digit PIN. Wait 15 minutes and try again with the correct PIN. If you have lost the PIN, you will need to re-run the extraction — PINs are generated fresh per job and are never stored anywhere we can retrieve them. All failure responses look identical on purpose, so an attacker cannot tell whether they got the PIN wrong, the token wrong, or guessed a non-existent job. **Where do I find my PIN after extraction finishes?** The 6-digit PIN is shown on screen immediately after you connect your org — it appears on the same page where your download link is displayed. Save it somewhere safe before you close the tab. PINs are generated once per extraction job using a cryptographically secure random number generator, and we do not store them anywhere retrievable. If you lose the PIN, the fastest path is to re-run extraction (each run is a fresh $15 job with a new PIN and new download link). **My download link stopped working. Why?** Download links expire after 48 hours and are capped at 5 successful downloads per job — whichever comes first. After either limit, the server permanently deletes the generated files and the link returns a 410 Gone. This is intentional: it minimizes the window during which a leaked link could be abused, and it keeps your metadata from sitting on our servers indefinitely. If you need to download again, re-run extraction to generate a new job with a fresh 48-hour window. **TrailMeta rejected my Salesforce login URL with a 400 error. What URLs are allowed?** For security (to prevent the server being tricked into connecting to arbitrary hosts), TrailMeta only accepts three shapes of login URL: (1) https://login.salesforce.com — standard production login. (2) https://test.salesforce.com — sandbox login. (3) https://.my.salesforce.com — MyDomain URLs (including sandbox variants like --.sandbox.my.salesforce.com). URLs must use HTTPS, port 443, and no path/query/fragment. If your org uses Enhanced Domains (mandatory since Salesforce Winter '24), your MyDomain URL will work. If you are seeing this error with a legitimate MyDomain URL, double-check you copied it correctly from Setup > My Domain. **I am getting "Too many requests" when trying to connect. What is going on?** The connect endpoint is rate-limited to 5 attempts per 15 minutes per IP address to block credential-stuffing attacks. If you triggered the limit by mistyping your password a few times, wait 15 minutes and try again. If you are behind a shared NAT (corporate network, VPN) and someone else on the same IP hit the limit, you may see this error without having attempted anything yourself — wait it out, or try from a different network. ## Key Pages - [Home](https://trailmeta.com/): Product overview, features, pricing, and FAQ - [Blog](https://trailmeta.com/blog): Guides, comparisons, and thought leadership on Salesforce + AI - [Help Center](https://trailmeta.com/help): 49+ articles covering setup, extraction, AI platform guides, security, pricing, and troubleshooting - [Security](https://trailmeta.com/security): Security policy — AES-256-GCM encryption, 48-hour auto-deletion, no record data - [Privacy Policy](https://trailmeta.com/privacy) - [Terms of Service](https://trailmeta.com/terms) ## Blog Posts ### [How to Feed Your Entire Salesforce Org to ChatGPT](https://trailmeta.com/blog/feed-salesforce-org-to-chatgpt) **Category:** Integration Guides · **Reading time:** 12 min Step-by-step guide to extracting your complete Salesforce org metadata and uploading it to ChatGPT, Custom GPTs, or ChatGPT Projects so the AI actually understands your objects, flows, Apex, and architecture. ### [How to Give Claude Full Context About Your Salesforce Org](https://trailmeta.com/blog/give-claude-salesforce-org-context) **Category:** Integration Guides · **Reading time:** 14 min Complete guide to extracting your Salesforce org metadata and loading it into Claude, Claude Projects, or Claude Code so it understands your objects, flows, Apex, triggers, and architecture. ### [Free Salesforce Skills Bundle for AI Coding Assistants (Claude, Cursor, Windsurf)](https://trailmeta.com/blog/free-salesforce-skills-bundle-ai) **Category:** Free Tools · **Reading time:** 7 min 350+ validated Salesforce skills for your AI coding assistant. Works with Claude Code, Cursor, Windsurf, GitHub Copilot, and Codex. Drop in your project root — your AI loads Salesforce best practices automatically. ### [The Missing Piece in AI for Salesforce: Why Context Beats Capability](https://trailmeta.com/blog/missing-piece-ai-salesforce-context) **Category:** Thought Leadership · **Reading time:** 10 min Every AI tool for Salesforce has the same blind spot: it does not understand YOUR org. Here is why org context, not model capability, is the bottleneck, and how to fix it. ### [AI-Ready Salesforce Documentation: What Every Admin Needs in 2026](https://trailmeta.com/blog/ai-ready-salesforce-documentation-2026) **Category:** Guides · **Reading time:** 11 min Traditional Salesforce documentation is built for humans. In 2026, your documentation needs to work for AI too. Here is what AI-ready org documentation looks like and why it matters. ### [Salesforce + Cursor IDE: Give Your AI Coding Assistant Full Org Context](https://trailmeta.com/blog/salesforce-cursor-ide-org-context) **Category:** Integration Guides · **Reading time:** 10 min Set up Cursor IDE with complete Salesforce org context. Objects, fields, flows, Apex, triggers, profiles. Your AI coding assistant finally understands your org. ### [Vibe Coding Salesforce: Why Your AI Needs Org Context First](https://trailmeta.com/blog/vibe-coding-salesforce-ai-context) **Category:** Guides · **Reading time:** 11 min Vibe coding is transforming software development. But vibe coding Salesforce without org context produces code that compiles and fails in production. Here is how to do it right. ### [Agentforce is Not Enough: The Org Context Gap Salesforce AI Cannot Fill](https://trailmeta.com/blog/salesforce-agentforce-context-gap) **Category:** Thought Leadership · **Reading time:** 9 min Agentforce handles in-org AI actions. But when you need to plan, audit, migrate, or develop with external AI tools, you hit a wall. Here is the gap and how to close it. ### [5 Export Formats for 13+ AI Platforms: How TrailMeta Matches Your Stack](https://trailmeta.com/blog/trailmeta-export-formats-ai-platforms) **Category:** Guides · **Reading time:** 11 min TrailMeta produces 5 export formats optimized for different AI platforms. Single File for chat, Consolidated Pack for projects, Full Pack for search, IDE Project for coding assistants, API Bundle for developers. Here is which format fits your stack and why it cuts your token costs. ### [MCP Connectors vs Full Org Context: Which Does Your AI Actually Need?](https://trailmeta.com/blog/salesforce-mcp-vs-full-org-context) **Category:** Thought Leadership · **Reading time:** 10 min Salesforce MCP connectors give AI live query access. TrailMeta gives AI complete org understanding. They solve different problems — and one burns far more tokens than the other. Here is when to use each. ### [How to Map Salesforce Technical Debt with AI](https://trailmeta.com/blog/salesforce-technical-debt-ai-mapping) **Category:** Guides · **Reading time:** 10 min Your Salesforce org has technical debt you cannot see from Setup. Here is how to extract your complete org metadata and use AI to identify unused components, redundant automation, security gaps, and architecture risks. ### [How to Document Your Salesforce Org: Every Method Compared](https://trailmeta.com/blog/how-to-document-salesforce-org) **Category:** Guides · **Reading time:** 12 min Complete guide to documenting a Salesforce org — from manual spreadsheets to automated extraction. Covers what to capture, tool comparison, and how to produce AI-ready org documentation in minutes. ### [Salesforce Metadata Export Tools Compared (2026)](https://trailmeta.com/blog/salesforce-metadata-export-tool) **Category:** Guides · **Reading time:** 11 min Compare every tool for exporting Salesforce metadata — Salesforce CLI, Workbench, Elements.cloud, Metazoa, and TrailMeta. Learn which APIs each uses, what they export, and which produces AI-ready output. ### [The Complete Salesforce Org Audit Checklist (2026)](https://trailmeta.com/blog/salesforce-org-audit-checklist) **Category:** Guides · **Reading time:** 13 min A 40+ item checklist for auditing your Salesforce org: data model, automation, code quality, security, integrations, and packages. Includes AI-assisted audit workflow to automate 80% of the work. ### [Salesforce Technical Debt Assessment: A Scoring Framework](https://trailmeta.com/blog/salesforce-technical-debt-assessment) **Category:** Guides · **Reading time:** 12 min A structured framework for assessing Salesforce technical debt: inventory, classification, severity scoring, and remediation planning. Includes a priority matrix and example assessment output.