# Research notes: Amoeba by last even

Status: draft-for-human-review. No human reviewed, edited or fact-checked this package.

Submission state: **complete-draft**. Every published trait reproduces; the unresolved items are properties of the project record rather than gaps in the research, and are recorded as editorial exceptions.

## Per-essay checks report

| Check | Report |
| --- | --- |
| Identity and package | Chain 1, contract `0xa7d8d9ef8d8ce8992df33d8b8cf4aebabd5bd270`, project ID 182, slug `amoeba-by-last-even`, title `Amoeba`, artist credit `last even` as the index returns it. Identity checked against the returned composite key `...-182`. The platform renders the artist name three ways and the package records all three. All seven required files present. |
| Structure and citations | 5,193 words in the seven canonical sections in the required order, generated into essay.md from `record.json.sections` by the same expression the supplied validator uses. Eight unique source IDs, all cited and all resolving both ways. Significance ends with six explicitly unresolved research questions. |
| Source research | Eight materially cited inspected sources with specific locators: collection page blocks by name and quoted clause; four GraphQL retrievals by operation, variables and returned field; program findings by function, table and expression; live observations by token, global variable value, measured pixel census and canvas dimension; the p5 reference by quoted line; licence terms by clause. Remaining access gaps: the artist website of record is dead and was recorded as a negative check with all URLs and times; the collection page accordions were not expanded; the CC legal code was not opened, only the deed; the p5.js licence was not established; no on-chain transaction values were read; the feature-extraction software was not located. |
| Code | Retrieved by HTTP POST, named operation `ProjectBySlug`, variables `{"slug":"amoeba-by-last-even"}`, at 2026-09-18T07:47:45.489Z. 3,259 bytes, single fragment, runtime `p5@1.0.0`. Digest `835b79ebfb2da2b3de118e6dbe02d4346b5f1b52301f5452abcf5510d26f3eaf`, identical in `record.json.codeInspection.sha256` and `evidence.json.codeSha256` because both are written from the same measured value. Full reading covered seed handling and its precision loss, the 1,000-value generator, the weighted tables, the five motion tables, the per-frame placement of 160 shapes, the wall-clock time base, canvas geometry, the keypress handler and the dependency. Execution disclosure: read from text without local execution, and separately executed by the official renderer during observations. No artist code reproduced. |
| Visual evidence | Four distinct token identities observed, V01 to V04, each with chain, contract, token ID, canonical URL, inspected renderer URL, observer, method, wall-clock window and concrete observations. Two sustained readings: V01, with all four controls exercised, and V02, with a full chromatic census and a direct reading of the undefined stroke-weight global. Observer is this agent throughout. All four observed with the browser pane displayed and the page's own animation loop running, so no manual frame stepping was used on this project. Controls: all four quality keys exercised in turn on V01 and verified against both the resolution scale and the canvas dimensions; the 's' save handler deliberately not exercised because it would initiate a download. Audio: auditioned negatively and established as silent from both the program and the running instance. |
| Header | Complete linked wallet set refreshed this session; three addresses, no fixed count assumed. All 213 token owner addresses compared against all three. One match: token 182000181. No collection gap. |
| History and rights | Four events kept distinct, and the finding that activation postdates the first mint by eight days is reported rather than smoothed. Sale duration given with both endpoints and the calculation, and paired with a mint-pace curve because a single figure is misleading for an edition whose last tenth took more than a year. The charity window is quantified against the mint curve. No sale mechanism is asserted. Rights domains kept separate, and the missing licence version is treated as the most consequential gap in the record rather than quietly filled in. |
| Provenance and usage | Original author recorded as Anthropic / claude-opus-5, routing label rejected as a model identifier. No revisions. Usage counts null with reasons; `startedAt` timezone-aware, `cutoff` null with explanation. |
| Validator | Command, identity, exit status and unedited output below. |
| Submission state | **complete-draft**. Validator passes; research requirements met; eleven editorial exceptions recorded. |

## Corrections made during research

None recorded for this package. One intermediate result was revised in the course of the work and is reported in its corrected form: the `harmony` trait initially appeared to mismatch on one token until the ghost override was identified, after which it reproduced at 213 of 213.

## Validator

Command actually executed, from the extracted pack root:

```sh
python3 scripts/validate_series_essays.py submission/amoeba-by-last-even
```

Supplied validator identity: `scripts/validate_series_essays.py` as distributed in this pack, SHA-256 `3b6273b6edaa9eee0f7641f365d4d243f15977e0acdd85bac922a40f157472ca`, 61 lines. No substitute validator was used. A structural pass is not by itself completion; the checks report above is the substantive report.

Exit status: `0`

Unedited output:

```
{
  "status": "PASS",
  "scope": "Structural package validation only. Does not grant editorial acceptance or confirm publication.",
  "essays": [
    {
      "title": "Amoeba",
      "slug": "amoeba-by-last-even",
      "words": 5193,
      "sources": 8,
      "examples": 4,
      "pendingEditorialExceptions": 11
    }
  ],
  "totalWords": 5193,
  "errors": []
}
```