Repository navigation
Make the per-compound calculators answer the whole query - #91
Merged
Merged
Conversation
The first version was a calculator with a dose table under it, about 750 words. It missed most of what people actually search for on these terms. The dominant intent is one question in different words: how many units is this dose, in this vial, with this much water. People want to read that off a chart rather than type it in, which is why every page ranking for these terms has one. These now do: a table per vial size of 2, 5 and 10 mg, columns for 1, 2 and 3 mL of diluent, cells in units on a U-100 syringe, with anything over 100 units marked because it will not fit in one syringe. The arithmetic was checked against published charts: a 2 mg vial in 1 mL gives 12.5 units for 0.25 mg, and in 2 mL gives 25, which is what the reference tables say. The rows are what make the chart belong to a compound rather than being the same grid nineteen times: they are the absolute doses that studies of that compound actually administered. Per-kilogram figures are deliberately excluded from the parser. For four compounds that leaves nothing to chart, because every dose ever published for them is per kilogram, and those pages now say so in a section of their own rather than quietly dropping a block. Multiplying a rat milligram-per-kilogram figure by a human body weight is the one operation this site must not imply, so the page names it and explains why interspecies scaling is not multiplication. Five more blocks come from cited data the records already held and the first version ignored: weight-normalised doses with the reason they are off the chart, escalation schedules where a trial titrated, how long studies ran, storage and stability as published, and routes. A FAQ answers the unit question in the searcher's own words and says where the numbers come from. Mean length goes from about 750 words to 1,562, and the additions are computed arithmetic or quoted evidence rather than filler. SourceIds.astro exists because Astro frontmatter is TypeScript, not JSX, so a helper returning markup cannot live up there. The tests now check each page either charts doses or explains why it cannot, never both, and that the FAQ and the refusal are present, since the refusal is what keeps these inside the site's scope rule. Verified by removing each of the three in turn. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This branch was successfully deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
You were right that the first version was too thin. It was a calculator with a dose table under it, about 750 words, and it missed most of what people search for on these terms.
What was missing
The dominant intent is one question asked in different words: how many units is this dose, in this vial, with this much water. People want to read that off a chart, not type it in, which is why every page ranking for these terms has one and mine did not.
Each page now carries a chart per vial size (2, 5 and 10 mg), with columns for 1, 2 and 3 mL of diluent and cells in units on a U-100 syringe. Anything over 100 units is marked, because it does not fit in one syringe.
The arithmetic was checked against published charts:
Which is exactly what the reference tables give.
The rows are what make a chart belong to a compound rather than being the same grid nineteen times: they are the absolute doses studies of that compound actually administered.
Per-kilogram doses are refused, and four pages say so
The parser deliberately skips
mg/kgfigures. For humanin, melanotan II, selank and TB-500 that leaves nothing to chart, because every dose ever published for them is per kilogram.Those four now carry a section explaining the absence instead of quietly dropping a block. Multiplying a rat per-kilogram figure by a human body weight is the one operation this site must not imply, so the page names it and explains that interspecies scaling is a calculation with published factors, not a multiplication.
Five more blocks, all from cited data the records already held
Weight-normalised doses with the reason they are off the chart, escalation schedules where a trial titrated, how long studies ran, storage and stability as published, and routes. Plus a FAQ that answers the unit question in the searcher's own words.
The additions are computed arithmetic or quoted evidence, not filler.
Tests
Each page must either chart doses or explain why it cannot, never both, and must carry the FAQ and the refusal, since the refusal is what keeps these inside the site's scope rule. Verified by removing each of the three in turn, all three fail the build.
One thing I could not do
peptidecalcs.comsits behind a Vercel bot checkpoint that requires JavaScript, so curl, WebFetch and the archive all return a challenge page rather than content, and no browser tool is available in this session. I built the intent model from search results and published reconstitution charts instead. If you save one of their pages to the repo I will compare directly and close any remaining gap.🤖 Generated with Claude Code