# How do I mint and look up species \(with pictures\)?

Checked 2026-11-15 against GilbApp d5296208bbffa2ab70076e6194f5537c4be1ddf7.

Search GilbLife with gilb-mint.ts search, send missing species with gilb-mint.ts mint, then send their pictures with gilb-mint.ts mint.

## Before you start

gilb-mint.ts signed in, as [the answer on minting a term](http://support.gilb.com/help/how-do-i-mint-a-term-from-code-the-terminal-kit-or-chat) shows. For pictures, each picture and a smaller copy of it [as files in your own vault](http://support.gilb.com/help/how-do-i-upload-download-and-end-files).

## Steps

1. Search GilbLife for the species first.

   Search by the species' GBIF address, or by its scientific or common name:
   node \~/.gilb/kit/gilb-mint.ts search &lt;the species' GBIF address\>

   Species are kept apart, in GilbLife, GilbPlatform's list of species. GBIF, the Global Biodiversity Information Facility, keeps a public list of species, and a species' GBIF address looks like https://www.gbif.org/species/&lt;number\>. Words are asked of GilbVeda first, then of GilbLife as a scientific or common name in any language, case ignored. The answer's @graph holds each species found, at most 20 for a name, with @id, its address in GilbLife; taxonID, its GBIF address; scientificName; taxonRank; status; termMintedBy; and its pictures. An empty @graph for a GBIF address means GilbLife does not have it; for a name it proves nothing, as the species may carry another name, or the answer says partial. A "MISS" in GilbVeda for a species' name is no reason to mint a term.
2. Send the species that are missing.

   Write { "taxa": \[ ... \] } with each species' scientificName and, if you have them, its names, at most 25 a request, and send it:
   node \~/.gilb/kit/gilb-mint.ts mint &lt;species file\>

   names is { "preferred": { "en": "&lt;one name\>" }, "local": { "nb": \["&lt;a name\>"\] } }: one preferred name and a list of other names for each language tag, a short code for a language, such as en for English. GilbPlatform finds each species in GBIF by an exact match on its scientific name, builds its line of ancestors up to Life, and writes each missing one once; a species already there is reused. Send nothing else: GBIF fills the rest. A 201 answer lists minted, ancestors included, and reused; 200 means nothing new. 422 "LifeChainIncomplete": no exact match or no whole line, and nothing was written. 503 "GbifUnavailable": GBIF did not answer, so send it again later.
3. Send their pictures.

   Write { "pictures": \[ ... \] } naming, for each picture, the species' taxonID, the files of the picture and of its copy in your own vault, the credit and the licence, and send it:
   node \~/.gilb/kit/gilb-mint.ts mint &lt;pictures file\>

   Each picture is { "taxonID": "&lt;GBIF address\>", "original": { "derivedFrom": "&lt;file address\>" }, "copy": { "derivedFrom": "&lt;file address\>" }, "attribution": "&lt;credit\>", "license": "CC BY 4.0", "main": true }. The program reads each file with your sign-in, so only files you may read are sent. The copy must be smaller and show the whole picture, same shape: a cut-out answers 422 "PictureNotWholeFrame". The licence is "CC BY 4.0" or "CC0 1.0"; main asks for this to be the main picture when the species has two or more. 201 lists stored, 200 means nothing new, and 422 "LifeTaxonNotFound" means the species must be sent first.

## Deeper

Also asked:

How do I look up or add species, with pictures? Search GilbLife with gilb-mint.ts search, send the missing species with gilb-mint.ts mint, then send their pictures with gilb-mint.ts mint, as the steps show.

What you send lands as "Suggested", with you as its minter, and Agent.Concept, the agent that finds and makes GilbVeda's terms, approves it later. The species and the pictures go through Agent.Concept as terms do: the program fetches the pass for GilbLife, and a request without it answers 403 "NotThroughAgentConcept". The search needs no pass, only a sign-in: with no token it answers 401 "This surface requires an authenticated caller \(no token presented\)". Every signed-in person is answered every species in GilbLife. Species and pictures count against the same daily mint limits as terms, counted apart for GilbLife.

One file of a picture is at most 2 MiB, and one request carries at most 8 pictures; when a file holds more species or pictures than one request takes, the server answers 413 "LifeMintBatchTooLarge" with the number, and the program sends the rest in pieces. Sending the same species or picture again writes nothing. Each answer says under mark what happened to the species' main picture, such as "written", "kept" or "missing". A JPEG, the usual kind of photo file, loses the place it was taken and the camera's serial numbers. The picture is public. A picture made by AI is under CC0 1.0, a licence that lets anyone use it with no credit owed; for your own real photo you choose CC BY 4.0, which asks for credit, or CC0 1.0; when it is not known which, it asks you whether AI made it. "All rights reserved" and any licence that blocks another app are refused.

While a species or picture you sent is still "Suggested", you may [change or retire it](http://support.gilb.com/help/how-do-i-change-or-retire-a-term-i-minted) with the same program's change and retire. A change file is { "taxonID", "names" } for a species' names, or { "@id": &lt;the picture's original\>, with attribution, license, or main together with "taxonID" } for a picture; GBIF's facts cannot be changed. A retire file is { "@id": &lt;your species, or your picture's original\>, "deprecationNote": "&lt;why\>" }. Retiring never deletes. A species is retired from the bottom up: not while a species below it, or a picture, still hangs on it, which answers 409 "LifeRecordInUse". 403 "NotYourLifeRecord", 409 "LifeRecordNotSuggested" and 409 "LifeRecordRetired" mean the same as for terms.

From your own code, the routes are on https://life.gilb.com: GET /base/life/taxa with taxonID or scientificName, and POST /base/life/mint, /base/life/mint/pictures, /base/life/mint/change and /base/life/mint/retire, each with the person's token and [a new DPoP proof](http://support.gilb.com/help/how-do-i-make-a-dpop-proof-for-each-request) \(a short signed note your app makes for one request, naming its method and address\); the four POST routes also need the pass from an Agent.Concept turn opened on that host; examples 1 and 2 show them. A picture sent this way carries each file's bytes as base64, the standard way of writing a file's bytes as text. In GilbApp's chat, Agent.Concept looks species up and adds them, but stores no picture; [the answer on setting a term's picture](http://support.gilb.com/help/how-do-i-set-a-terms-picture) shows GilbApp's way, and [the answer on browsing the tree of life](http://support.gilb.com/help/how-do-i-browse-the-tree-of-life-species) shows how to see them. Minting species is not in the Developer Guide, GilbPlatform's guide for people who build apps on it, yet.

Source: GilbPlatform 498d29a0 \(live at https://id.gilb.com/version, read 2026-11-11\): src/routes/life-mint.ts:1-36 \(routes; same gate; own records only while Suggested; lookup a signed-in read\), :61-67 \(LIFE\_TARGET\), :76-102 \(pictures body, request-level license and attribution, change and retire bodies\), :137-178 \(credit and licence required\), :197-212 \(refusal names\), :214-230 \(MintQuotaError, GbifUnavailable 503, LifeStoreMoved 409\), :243-272 \(201 minted/reused, 200 nothing new\), :324-418 \(routes; MAX\_PICTURES\_PER\_REQUEST too large 413\), :428-446 \(GET /base/life/taxa: requireVerifiedCaller, then the query, then the host\); src/gilblife/answers.ts:30-58 \(exactly one of taxonID, scientificName\), :63-157 \(species keys, names shape, language tags, SPECIES\_NOT\_KEYS\), :164-186 \(413 LifeMintBatchTooLarge with maximum; MAX\_SPECIES\_PER\_REQUEST\), :199-266 \(taxonRow; MAX\_TAXA\_BY\_NAME 20; every live record answered to any verified caller; partial note\); src/gilblife/mint.ts:81 \(MAX\_SPECIES\_PER\_REQUEST 25\), :478-479 \(the same quotas, counted in the life vault\); src/gilblife/picture-mint.ts:1-34 \(two files, credit and licence on the original, resend writes nothing\), :64 \(8 a request\), :67 \(2 MiB a file\), :80-166 \(whole-frame check\), :120-140 \(JPEG GPS and serials stripped\), :290-366 \(mark words notNeeded, written, moved, kept, notMoved, writtenForExisting, missing\); src/gilblife/picture-license.ts \(CC BY 4.0, CC0 1.0 only\); src/gilblife/own-records.ts:1-26 \(change names or a picture's credit, licence, main; retire leaf-first, no cascade\), :63-67, :268-282 \(messages\). src/routes/\_mint-gate.ts:1-42 \(one gate for terms and GilbLife: host 404, mayMintTerms 401/403, vault served and local 503, pass 403 NotThroughAgentConcept "New ontology is minted through Agent.Concept only. ..."\), :92-128; src/auth/mint-pass.ts:1-66 \(pass issued by the server for an Agent.Concept turn; header x-base-mint-pass; aud base:mint:vocab or base:mint:life; 15 minutes; sub = the person\); src/routes/ai-claude-code.ts:271-298 \(POST /base/ai/claude-code/context: mintPasses only on an Agent.Concept turn and when mayMintTerms passes\), :114-127 \(body: userText required, agent, agentEntry, contentVaultIds, turnId\); src/auth/open-mint-standing.ts:1-110 \(a signed-in person or a bound AI; computer credential or unbound machine 403 NotAPersonToMint; no http principal 401 SignInToMint\). scripts/devkit/gilb-mint.ts:41-45 \(id.gilb.com, veda.gilb.com, life.gilb.com\), :186-239 \(usage text\), :271-317 \(signin: request-code; code: verify-code with the kit's own P-256 key; \~/.gilb/mint-credential.json 0600; expiry check\), :324-357 \(passFor: POST &lt;base\>/base/ai/claude-code/context with agent Agent.Concept, mintPasses.vocab\|life; needsPass for POST /base/vocab/, /base/life/mint, /base/term-requests\), :502-540 \(search: one https IRI = GilbLife taxonID; words = term-requests then GilbLife scientificName; --app prints "usage: recorded for" / "usage: NOT recorded for"\), :542-581 \(mint/change/retire routing; the shipped example refused locally, exit 2\); scripts/devkit/README.md:21-36 \(Node 22.18+ or 23.6+; 6-digit code 10 minutes; credential 7 days\), :38-83 \(the loop, answers, exit codes 0/1/2, dropped connection\), :85-143 \(GilbLife\), :171-184 \(a member\); scripts/devkit/example-package.json \(an attribute and its Value\). Agent.Concept \(J 06gf80rjtdw3h9q187p80kx00m, 2026-11-11\): Task \(any signed-in person; terminal: node \~/.gilb/kit/gilb-mint.ts, which the one setup link installs; look up by name and by meaning; HIT = use it; MISS = mint at once; usage row; no record stands in\), ClosedSend, PictureLicence. Public reads, no token, 2026-11-11: https://life.gilb.com/base/life/taxa?taxonID=&lt;a GBIF address\> and ?scientificName=&lt;a name\> 401 "This surface requires an authenticated caller \(no token presented\)". Shared sentences, word for word: species, the licence sentence, JPEG.

## For developers

### curl: species.sh

```curl
# 1. Open an Agent.Concept turn on GilbLife's host. Keep mintPasses.life from the answer.
curl -s https://life.gilb.com/base/ai/claude-code/context \
  -H "Authorization: DPoP <sign-in token>" \
  -H "DPoP: <a new proof for POST https://life.gilb.com/base/ai/claude-code/context>" \
  -H "Content-Type: application/json" \
  --data '{"userText":"<which species, and why>","agent":"Agent.Concept"}'

# 2. Is the species in GilbLife? By its GBIF address, encoded (or scientificName=<name>).
curl -s "https://life.gilb.com/base/life/taxa?taxonID=<GBIF address, encoded>" \
  -H "Authorization: DPoP <sign-in token>" \
  -H "DPoP: <a new proof for GET https://life.gilb.com/base/life/taxa>"
# {"@graph":[]} for a GBIF address: not in GilbLife

# 3. Send it. GBIF fills in the rest.
curl -s https://life.gilb.com/base/life/mint \
  -H "Authorization: DPoP <sign-in token>" \
  -H "DPoP: <a new proof for POST https://life.gilb.com/base/life/mint>" \
  -H "X-Base-Mint-Pass: <mintPasses.life>" \
  -H "Content-Type: application/json" \
  --data '{"taxa":[{"scientificName":"<scientific name>","names":{"preferred":{"en":"<one name>"}}}]}'
# 201: {"minted":[{"@id":"<address>","taxonID":"<GBIF address>","scientificName":"<scientific name>","taxonRank":"<rank>","status":"Suggested"}, ...],"reused":[...],"termMintedBy":"<your address>"}
```

### curl: pictures.sh

```curl
# Send a picture and its smaller, whole-frame copy, each as base64, with the credit and the licence.
curl -s https://life.gilb.com/base/life/mint/pictures \
  -H "Authorization: DPoP <sign-in token>" \
  -H "DPoP: <a new proof for POST https://life.gilb.com/base/life/mint/pictures>" \
  -H "X-Base-Mint-Pass: <mintPasses.life>" \
  -H "Content-Type: application/json" \
  --data @pictures.json
# pictures.json: {"pictures":[{"taxonID":"<GBIF address>",
#   "original":{"bytes":"<base64>","derivedFrom":"<the file's address in your vault>"},
#   "copy":{"bytes":"<base64>","derivedFrom":"<the copy's address in your vault>"},
#   "attribution":"<the credit>","license":"CC BY 4.0","main":true}]}
# 201: {"stored":[{"taxonID":"<GBIF address>","original":{...},"copy":{...},"status":"Suggested","mark":"<what happened to the main picture>"}],"unchanged":[],"termMintedBy":"<your address>"}
```
