From e4b17febd7a047ffcac74d09ed810ea2263b9dfe Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Pedro=20Hern=C3=A1ndez?= Date: Mon, 7 Sep 2026 15:35:29 -0400 Subject: [PATCH] =?UTF-8?q?0.7.1=20=E2=80=94=20the=20version=20it=20report?= =?UTF-8?q?s,=20and=20the=20page=20teachers=20land=20on?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Two corrections that never reached anyone from 0.7.0, both about what this project says about itself rather than what it measures. `signsofai --version` reported 0.1.0 for seven releases (#84). It now reads the assembly, and this release is the first time that is visible: the tool built from these files answers 0.7.1, which is what the csproj packs. `why.html` — the page README line 15 sends teachers to first — was still showing ninety texts and a green zero six days after the published rate stopped being zero (#85). It now carries 296, two flagged, a bound of 2.4%, and the paragraph that inverted: 206 of the corpus are classroom essays by adults learning English, which is the group the boundary moved to protect. The guard added with that page earned its keep immediately. Bumping the version made the calibration declare 0.7.1, and WhyPageTests failed by name — `The_page_names_the_engine_that_produced_its_figures` — instead of letting the page go stale in silence the way it did in August. That is the whole argument for a guard over a comment, demonstrated one day later. Calibration re-run: only the `engine` field moved. Threshold 30, 2 of 296, both bounds, every language and group row, the corpus fingerprint — unchanged. No score changes in this release; `Core`'s only edits are comments. `DesktopRelease.Version` is 0.7.1, so `desktop-v0.7.1` has to be cut from this branch before it merges, or the download button points at a 404. 432 tests. Co-Authored-By: Claude Opus 5 Claude-Session: https://claude.ai/code/session_015PEbbiYSNPw7jE3LrPNhyF --- Docs/CALIBRATION.md | 2 +- src/SignsOfAI.Cli/SignsOfAI.Cli.csproj | 2 +- src/SignsOfAI.Core/Calibration/published-calibration.json | 2 +- src/SignsOfAI.Core/SignsOfAI.Core.csproj | 2 +- src/SignsOfAI.Mcp/.mcp/server.json | 4 ++-- src/SignsOfAI.Mcp/SignsOfAI.Mcp.csproj | 2 +- src/SignsOfAI.UI/Services/DesktopRelease.cs | 2 +- src/SignsOfAI.Web/wwwroot/why.html | 6 +++--- 8 files changed, 11 insertions(+), 11 deletions(-) diff --git a/Docs/CALIBRATION.md b/Docs/CALIBRATION.md index 4521717..868297a 100644 --- a/Docs/CALIBRATION.md +++ b/Docs/CALIBRATION.md @@ -11,7 +11,7 @@ It is **not an accuracy figure**. Accuracy needs machine-written text to measure - **Corpus** `signsofai-human-baseline`, fingerprint `78bda061bde3dc99` - **Texts** 296 (452,184 words) - **Lengths measured** 649 – 9,328 words (median 832) -- **Engine** SignsOfAI.Core 0.7.0 +- **Engine** SignsOfAI.Core 0.7.1 - **Run** 2026-09-07 - **Target false-positive rate** 5% diff --git a/src/SignsOfAI.Cli/SignsOfAI.Cli.csproj b/src/SignsOfAI.Cli/SignsOfAI.Cli.csproj index f0adf1f..009d0fb 100644 --- a/src/SignsOfAI.Cli/SignsOfAI.Cli.csproj +++ b/src/SignsOfAI.Cli/SignsOfAI.Cli.csproj @@ -13,7 +13,7 @@ true signsofai SignsOfAI.Cli - 0.7.0 + 0.7.1 Pedro Hernández (PeopleWorks) PeopleWorks CLI to detect the signs of AI writing (English & Spanish) and recommend fixes — lint your prose in CI. diff --git a/src/SignsOfAI.Core/Calibration/published-calibration.json b/src/SignsOfAI.Core/Calibration/published-calibration.json index 7269ccd..f9d457b 100644 --- a/src/SignsOfAI.Core/Calibration/published-calibration.json +++ b/src/SignsOfAI.Core/Calibration/published-calibration.json @@ -1 +1 @@ -{"corpusId":"signsofai-human-baseline","texts":296,"measuredOn":"2026-09-07","engine":"0.7.0","recommendedThreshold":30,"flaggedAtThreshold":2,"rateLow":0.0018549062642001807,"rateHigh":0.024297110573472523,"shortestWords":649,"longestWords":9328,"noisiestRules":[{"ruleId":"lex.just","textShare":0.34797297297297297},{"ruleId":"stat.burstiness","textShare":0.32094594594594594},{"ruleId":"rhet.in-conclusion","textShare":0.30067567567567566},{"ruleId":"rhet.rule-of-three","textShare":0.26013513513513514},{"ruleId":"lex.moreover","textShare":0.22297297297297297},{"ruleId":"rhet.not-only-but","textShare":0.21283783783783783},{"ruleId":"lex.furthermore","textShare":0.1858108108108108},{"ruleId":"lex.crucial","textShare":0.16891891891891891}],"languages":[{"language":"en","texts":271,"recommendedThreshold":30,"bestBound":0.01397699931055571,"rateHighAtThreshold":0.026504623306870336},{"language":"es","texts":25,"recommendedThreshold":null,"bestBound":0.13319225276039096,"rateHighAtThreshold":null}]} +{"corpusId":"signsofai-human-baseline","texts":296,"measuredOn":"2026-09-07","engine":"0.7.1","recommendedThreshold":30,"flaggedAtThreshold":2,"rateLow":0.0018549062642001807,"rateHigh":0.024297110573472523,"shortestWords":649,"longestWords":9328,"noisiestRules":[{"ruleId":"lex.just","textShare":0.34797297297297297},{"ruleId":"stat.burstiness","textShare":0.32094594594594594},{"ruleId":"rhet.in-conclusion","textShare":0.30067567567567566},{"ruleId":"rhet.rule-of-three","textShare":0.26013513513513514},{"ruleId":"lex.moreover","textShare":0.22297297297297297},{"ruleId":"rhet.not-only-but","textShare":0.21283783783783783},{"ruleId":"lex.furthermore","textShare":0.1858108108108108},{"ruleId":"lex.crucial","textShare":0.16891891891891891}],"languages":[{"language":"en","texts":271,"recommendedThreshold":30,"bestBound":0.01397699931055571,"rateHighAtThreshold":0.026504623306870336},{"language":"es","texts":25,"recommendedThreshold":null,"bestBound":0.13319225276039096,"rateHighAtThreshold":null}]} diff --git a/src/SignsOfAI.Core/SignsOfAI.Core.csproj b/src/SignsOfAI.Core/SignsOfAI.Core.csproj index 7baa5be..427a7e2 100644 --- a/src/SignsOfAI.Core/SignsOfAI.Core.csproj +++ b/src/SignsOfAI.Core/SignsOfAI.Core.csproj @@ -10,7 +10,7 @@ SignsOfAI.Core - 0.7.0 + 0.7.1 Pedro Hernández (PeopleWorks) PeopleWorks The engine behind Signs of AI Writing: detects the stylometric tells of AI-generated text (English & Spanish) — overused vocabulary, rhetorical crutches, syntactic tells and burstiness — and returns an actionable fix for each finding. No external dependencies. diff --git a/src/SignsOfAI.Mcp/.mcp/server.json b/src/SignsOfAI.Mcp/.mcp/server.json index 95cc8be..9fba6f8 100644 --- a/src/SignsOfAI.Mcp/.mcp/server.json +++ b/src/SignsOfAI.Mcp/.mcp/server.json @@ -1,7 +1,7 @@ { "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json", "name": "io.github.peopleworks/signs-of-ai", - "version": "0.7.0", + "version": "0.7.1", "title": "Signs of AI Writing", "description": "AI-writing detection (EN/ES) that shows the evidence: named tells, hidden characters, citations.", "packages": [ @@ -9,7 +9,7 @@ "registryType": "nuget", "registryBaseUrl": "https://api.nuget.org/v3/index.json", "identifier": "SignsOfAI.Mcp", - "version": "0.7.0", + "version": "0.7.1", "transport": { "type": "stdio" }, diff --git a/src/SignsOfAI.Mcp/SignsOfAI.Mcp.csproj b/src/SignsOfAI.Mcp/SignsOfAI.Mcp.csproj index 8c89780..00fcec7 100644 --- a/src/SignsOfAI.Mcp/SignsOfAI.Mcp.csproj +++ b/src/SignsOfAI.Mcp/SignsOfAI.Mcp.csproj @@ -12,7 +12,7 @@ true signsofai-mcp SignsOfAI.Mcp - 0.7.0 + 0.7.1 Pedro Hernández (PeopleWorks) PeopleWorks Model Context Protocol (MCP) server for Signs of AI Writing. Exposes the engine as tools — analyze text for AI tells, check originality between documents, search the sign catalog, extract distinctive phrases, and (optionally, via the hosted API) measure perplexity and find paraphrases — to Claude Desktop and any MCP client. diff --git a/src/SignsOfAI.UI/Services/DesktopRelease.cs b/src/SignsOfAI.UI/Services/DesktopRelease.cs index 4d83c76..b589964 100644 --- a/src/SignsOfAI.UI/Services/DesktopRelease.cs +++ b/src/SignsOfAI.UI/Services/DesktopRelease.cs @@ -20,7 +20,7 @@ namespace SignsOfAI.UI.Services; public static class DesktopRelease { /// The published version, without the desktop-v prefix its tag carries. - public const string Version = "0.7.0"; + public const string Version = "0.7.1"; /// /// The .zip itself, so the button downloads rather than starting a scavenger hunt through a diff --git a/src/SignsOfAI.Web/wwwroot/why.html b/src/SignsOfAI.Web/wwwroot/why.html index 5659b25..30c23dd 100644 --- a/src/SignsOfAI.Web/wwwroot/why.html +++ b/src/SignsOfAI.Web/wwwroot/why.html @@ -23,7 +23,7 @@ Spanish is written in Spanish rather than translated into it, as everything else here is. Every figure comes from src/SignsOfAI.Core/Calibration/published-calibration.json, measured on - 2026-09-07 with engine 0.7.0. If that file changes, this page is stale and must be regenerated + 2026-09-07 with engine 0.7.1. If that file changes, this page is stale and must be regenerated by hand — there is no build step behind it, deliberately: it has to open from a downloads folder years from now on a machine that has never heard of this project. @@ -549,12 +549,12 @@

What do I do on Monday?