Peptide Portal Compound Research

Purpose

Given a peptide name (e.g. BPC-157, tirzepatide), an alias (Body Protection Compound), a CAS number (137525-51-0), or a use-case keyword that maps to Peptide Portal's category taxonomy (GLP-1, GH secretagogue, tissue repair), return a structured profile aggregating:

• Vendor pricing — the aggregated $ / mg range and vendor count (public), or the per-vendor catalog rows (auth-gated, see Gotchas).
• Purity / COA signal — the marketplace's Verified / Unverified COA flag and the vendor's third-party-testing percentage.
• Research guide — the deep-link /peptides/{slug} clinical brief (FDA status, evidence tables, comparison with similar compounds) and any /blog/{slug} topical guides that cite dosing data.
• Clinical paper references — the per-peptide bibliography at /research/{slug}, where each paper has a dedicated detail page /research/{slug}/{uuid} linking out to PubMed.

Read-only — never click CREATE FREE ACCOUNT, Sign in, or GET ACCESS. Never submit the newsletter form.

When to Use

• Build a research dossier on one peptide before sourcing.
• Compare two peptides (e.g. BPC-157 vs TB-500) — fetch both and diff the structured outputs.
• Surface the canonical published evidence index for a peptide (citations, study designs, models, indications) without re-scraping PubMed yourself — Peptide Portal already curates it per compound.
• Find which peptides on the catalog match a use-case keyword (GLP-1, growth hormone, tissue repair, cognitive) and rank by COA-verified vendor count.

Workflow

Peptide Portal is a public Vercel/Next.js site with no anti-bot, no captcha, no Akamai gating for read paths. a browserless_agent session with default flags works fine; a stealth + residential-proxy session is not required for this site. All read URLs return server-rendered HTML (/api/* is blocked by robots.txt and returns 404, so there is no public JSON API to short-circuit to — the recommended method is browser).

There are four URL surfaces. Use them in this order:

1. Resolve the query to a peptide slug

If the user gave a known name, alias, or CAS:

• Direct slug guess (fastest path — ~150ms): https://peptideportal.org/peptides/{kebab-name}. The catalog has 58+ slugs in the sitemap (/sitemap.xml). Examples: bpc-157, tb-500, tirzepatide, semaglutide, cjc-1295, ipamorelin, ghk-cu, mots-c, ll-37, kpv, selank, semax, epitalon. A 200 response means the slug is canonical; a 404 means try the global typeahead.

• Global Cmd+K typeahead (fuzzy resolver — handles aliases, CAS, and vendor product names): on any page, click button: Search peptides, vendors, or CAS numbers. Press Command K to open. to open the dialog, then fill the combobox inside dialog: Search with the query. After ~1s the listbox under dialog: Search populates with up to three groups:

  Group label	What it contains
  Peptides	Catalog hits — peptide name + tagline + CAS. Click → /peptides/{slug}.
  Products	Per-vendor SKUs — name + vendor · peptide · category. Click → marketplace row.
  Actions	A Search marketplace for "{q}" fallback that navigates to /marketplace?q={q}.

  The typeahead substring-matches against peptide name, tagline (which contains use-case-ish phrases like "GLP-1 agonist", "GH secretagogue", "GIP / GLP-1 dual agonist"), CAS, and vendor product names. It does not map free-form use-case phrases — weight loss returns zero hits; the technical keyword glp returns Survodutide, Semaglutide, Tirzepatide cleanly. Lead with the canonical mechanism keyword, not the consumer phrase.

If the user gave a use-case keyword and the typeahead returns nothing, fall through to step 4 (browse the marketplace listing by category column).

2. Pull the research brief — /peptides/{slug}

goto https://peptideportal.org/peptides/{slug} then a snapshot. This is the canonical clinical brief and contains, in this fixed order, every section you need for the "research guide" output:

The detail page does not list per-vendor pricing inline. Vendor pricing lives on /marketplace and /vendor/{slug} (step 3).

3. Pull vendor pricing + purity — /marketplace?q={name}

goto https://peptideportal.org/marketplace?q={name} then a snapshot. The result is a one-row table with columns PEPTIDE | CATEGORY | RESEARCH FOCUS | $ / MG RANGE | VENDORS | COA | EXPAND. The single matching row exposes, without authentication:

• $ / mg range — aggregated min–max (e.g. $5.00–$8.00 for BPC-157).
• Vendors — integer count of vendors carrying this peptide.
• COA — Verified (third-party tested across vendor mix) or Unverified.
• Category — short label (e.g. Body Protective Compound).
• Research Focus — taxonomy bucket (e.g. Tissue Repair Research).

Per-vendor breakdown is auth-gated. Clicking the Expand button on the row reveals a row whose only content is View {Name} vendors — Create a free account to compare pricing across verified vendors — CREATE FREE ACCOUNT | Sign in. Do not click. Report the aggregated row and stop; record vendors_per_vendor_breakdown: "auth_required" in the output schema.

For an individual vendor's profile (vendor score, catalog breadth %, third-party-testing %, established date, visible category bucket list — all public): https://peptideportal.org/vendor/{vendor-slug}. Vendor slugs are discoverable from /vendors (e.g. peptide-crafters, peptidology, peptide-partners, orbitrex-peptides, skye-peptides, alchemy-peptides, nulife-peptides, risynth-bio, biotech-peptides, peptime, eternal-peptides, atomik-labz, kerisite-peptide, zlz-peptide). Detail pricing and COA documents on vendor pages are also auth-gated — only the scorecard summary is public.

4. Pull the bibliography — /research/{slug} and /research/{slug}/{uuid}

a residential-proxy HTTP fetch works fine here (no JS needed — the bibliography list is server-rendered). Each paper is a card link with text shaped like:

{Author} {Year} — {Title}
{Citations} {Journal} · {Year} {Study type} {Model} {Topic} Source {Tag}

…where {Study type} is one of Preclinical | Pilot | Review | Animal RCT | Systematic Review | Case Report | Safety, {Model} is Cell | Animal | Rat | Human | In Vitro | Various, and the href is /research/{slug}/{uuid} (UUID v4). Hit the per-paper UUID page for: outbound PubMed/DOI link (under heading External Links), full structured details (Study Design, Indication, Intervention, Species, Risk of Bias Assessment, Tags), citation count, evidence-quality grade, and last-verified date.

Not every peptide has a /research/{slug} index — peptides without curated evidence return 200 with an empty list, not a 404. Fall back to the Sources list at the bottom of /peptides/{slug} (step 2) which is always populated and links directly to FDA / PubMed / USADA / WADA / OPSS / journal pages.

5. Pull dosing-protocol guides — /blog/{slug}

Peptide Portal deliberately does not publish prescriptive dosing on /peptides/{slug} for unapproved compounds — for BPC-157, TB-500, etc., the detail page repeatedly says "human dosing is not established." Cited dosing data lives on long-form guides under /blog, where the doses are quoted from specific clinical studies (e.g. the CJC-1295 + Ipamorelin guide cites "30 or 60 μg/kg" and "0.03 mg/kg twice daily" with study references).

Discover blog guides matching the peptide via the sitemap (/sitemap.xml lists all 35 blog slugs) or via a substring filter on slug names (bpc-157, tb-500, cjc-1295, ipamorelin, glp1, peptides-for-hair-growth, etc.). Fetch via a direct HTTP fetch https://peptideportal.org/blog/{slug} — these are SSR HTML, no JS rendering required.

6. Release the session

browserless_agent sessions update "$the session" --status session-ends-on-return

Site-Specific Gotchas

• Auth wall on per-vendor pricing. The marketplace shows aggregated $ / mg range + vendor count + COA-verified flag publicly, but clicking Expand on any row reveals only a CREATE FREE ACCOUNT / Sign in paywall. Per-vendor SKU pricing, COA documents, and detailed vendor catalogs require an account. Report the aggregated public data and set a flag like per_vendor_pricing_visible: false in the output.
• Unauthenticated marketplace home is heavily curated. /marketplace (no q=) returns only 13 products even though /sitemap.xml lists 58+ peptide slugs. The 13 are the most-trafficked GLP-1s, tissue-repair, and GH peptides. To verify a specific peptide is in the marketplace, hit /marketplace?q={name} directly — it works for the full catalog (not just the 13 featured rows) provided the peptide has any vendor coverage.
• Cmd+K typeahead substring-matches name, tagline, CAS, and product names — but not free-form use cases. glp works (returns Survodutide, Semaglutide, Tirzepatide via the "GLP-1 agonist" and "GLP-1/glucagon dual agonist" taglines). weight loss returns zero peptide hits — only the Search marketplace for "weight loss" action fallback. Translate consumer phrases to mechanism keywords before submitting: weight loss → glp, obesity → glp, muscle gain → igf / ghrh, healing → bpc / tb-500, cognitive → selank / semax.
• /peptides page-local searchbox is name-only. The searchbox: Search peptide guides by peptide name on the /peptides index filters cards by substring of peptide name only — does not match aliases, CAS, taglines, or categories. Use the global Cmd+K dialog for anything beyond name.
• /peptides/{slug} detail page has no inline vendor pricing. Two link: Vendors anchors on the page navigate to the global /vendors page, not a per-peptide vendor block. Always pair /peptides/{slug} (research) with /marketplace?q={name} (pricing) — they are separate surfaces.
• Dosing is intentionally absent from /peptides/{slug} for unapproved compounds. Don't bother grepping the detail page for mg/kg — the page will say "human dosing is unclear." Cited doses live in /blog/{slug} guides instead, and only for compounds where clinical studies exist (CJC-1295 + Ipamorelin, semaglutide/tirzepatide titration, etc.). For unapproved peptides, the honest output is dosing_protocol: "no_human_protocol_published".
• /research/{slug} exists for every peptide but is sometimes empty. A 200 response with an empty card list means "no curated papers" — not "peptide not found." Treat empty as a normal outcome and fall back to the Sources block on /peptides/{slug} (which is always populated with regulatory links + at least one PubMed reference for in-scope peptides).
• Paper detail UUIDs are stable but per-peptide-namespaced. The same paper would have different UUIDs if it appeared under two peptides' research pages. Do not reuse a UUID across peptides; always re-derive from the parent /research/{slug} index.
• External paper links are PubMed-only on /research/{slug}/{uuid}. The detail page exposes a single outbound link: PubMed under heading External Links. No DOIs, no direct journal URLs. If the user needs a DOI, scrape PubMed by ID after this skill returns.
• /api/* is robots-disallowed and returns the SPA shell at 404. There is no JSON shortcut for any of these surfaces — the recommended method is unambiguously browser. a direct HTTP fetch (HTTP-only path) works for /peptides/{slug}, /research/{slug}, /research/{slug}/{uuid}, /vendor/{slug}, /blog/{slug}, /sitemap.xml, and /marketplace?q={name} (these are SSR HTML) — use it instead of a full browser session when you just need to read text. The pages that require an active browser are the global Cmd+K dialog and the marketplace filter sidebar (their popups are JS-driven).
• The "Compare verified vendors in one place" modal layers over /marketplace. It is a paywall promo that does not dismiss without auth — but the underlying table is rendered behind it and a snapshot traverses through. Don't waste turns trying to close it; snapshot still surfaces the row data.
• No CAS-number sanitization needed. CAS is exposed as a plain string with hyphens (137525-51-0) — Cmd+K matches it with or without spaces.
• Sitemap is your fastest enumeration tool. https://peptideportal.org/sitemap.xml lists every peptide slug, vendor slug, blog post, and research index, all under <loc> tags. Parse it once at the start of a bulk job instead of crawling.

Expected Output

A successful run returns one structured object per peptide. Three outcome shapes:

// 1. Full hit — peptide is in the catalog, marketplace, AND research index
{
  "success": true,
  "query": "BPC-157",
  "slug": "bpc-157",
  "name": "BPC-157",
  "aliases": ["Body Protection Compound-157", "PL 14736", "stable gastric pentadecapeptide BPC 157"],
  "cas": "137525-51-0",
  "category": "Body Protective Compound",
  "research_focus": "Tissue Repair Research",
  "fda_status": "not_approved",
  "legal_status": "research_use_only",
  "banned_in_sports": true,
  "research_brief_url": "https://peptideportal.org/peptides/bpc-157",
  "use_cases": [
    {"use": "Tendon healing", "evidence_level": "Preclinical", "known": "Animal and cell studies suggest effects on tendon fibroblast migration, survival, and healing pathways.", "not_known": "Whether it reliably improves tendon healing in humans is not established."},
    {"use": "Gut protection", "evidence_level": "Preclinical", "known": "Studied in gastrointestinal lesion and gastric integrity models.", "not_known": "Human benefit is unclear."}
  ],
  "vendor_pricing": {
    "marketplace_url": "https://peptideportal.org/marketplace?q=BPC-157",
    "price_per_mg_range_usd": {"min": 5.00, "max": 8.00},
    "vendor_count": 4,
    "coa_status": "Verified",
    "per_vendor_pricing_visible": false,
    "per_vendor_pricing_gate": "auth_required"
  },
  "dosing_protocol": "no_human_protocol_published",
  "related_guides": [
    {"title": "BPC-157 vs TB-500 Recovery Comparison", "url": "https://peptideportal.org/blog/bpc-157-vs-tb-500-recovery-comparison"},
    {"title": "FDA PCAC Meeting July 2026 — BPC-157, TB-500", "url": "https://peptideportal.org/blog/fda-pcac-meeting-july-2026-bpc-157-tb-500"}
  ],
  "clinical_papers": {
    "research_index_url": "https://peptideportal.org/research/bpc-157",
    "total_papers": 14,
    "papers": [
      {
        "title": "Chang 2011 — BPC-157 Promotes Tendon Fibroblast Outgrowth and Migration",
        "detail_url": "https://peptideportal.org/research/bpc-157/82802990-98f4-4ce2-b6d3-75702b44ded4",
        "pubmed_url": "https://pubmed.ncbi.nlm.nih.gov/21030672",
        "journal": "Journal of Applied Physiology",
        "year": 2011,
        "citations": 202,
        "study_type": "Preclinical",
        "model": "Cell",
        "indication": "Tendon healing mechanisms"
      }
    ]
  },
  "regulatory_sources": [
    {"title": "FDA: Certain Bulk Drug Substances for Use in Compounding May Present Significant Safety Risks", "url": "https://www.fda.gov/drugs/human-drug-compounding/certain-bulk-drug-substances-use-compounding-may-present-significant-safety-risks"},
    {"title": "USADA: BPC-157, A Prohibited Peptide", "url": "https://www.usada.org/..."},
    {"title": "WADA: 2022 Prohibited List Now in Force", "url": "https://www.wada-ama.org/..."}
  ]
}

// 2. Catalog hit, no marketplace coverage — peptide exists on /peptides/{slug} but not in /marketplace
{
  "success": true,
  "query": "epitalon",
  "slug": "epitalon",
  "name": "Epitalon",
  "research_brief_url": "https://peptideportal.org/peptides/epitalon",
  "vendor_pricing": {
    "marketplace_url": "https://peptideportal.org/marketplace?q=Epitalon",
    "price_per_mg_range_usd": null,
    "vendor_count": 0,
    "coa_status": null,
    "per_vendor_pricing_visible": false,
    "per_vendor_pricing_gate": "no_vendor_coverage"
  },
  "dosing_protocol": "no_human_protocol_published",
  "clinical_papers": {
    "research_index_url": "https://peptideportal.org/research/epitalon",
    "total_papers": 0,
    "papers": []
  }
}

// 3. Not found — query did not resolve to any catalog slug or typeahead hit
{
  "success": false,
  "reason": "peptide_not_found",
  "query": "made-up-peptide-xyz",
  "tried": [
    "https://peptideportal.org/peptides/made-up-peptide-xyz (404)",
    "global Cmd+K typeahead (0 Peptides group results)"
  ],
  "suggestion": "Try a canonical alias, CAS number, or mechanism keyword (e.g., 'GLP-1', 'GH secretagogue', 'tissue repair'). The full catalog of ~58 slugs is at https://peptideportal.org/sitemap.xml."
}