imbad0202__academic-research-skills/plugin-evals-citation-check/README.md

plugin-evals-citation-check — academic-paper citation-check flow

Eval suite for claude plugin eval. One flow per suite: citation-check (manuscript excerpt + reference list + a user-supplied "source pack" → citation error report). Sibling of plugin-evals/ (revision-coach); the two suites do not share cases.

Quality spec (author-defined, 2026-09-13): the four citation failures that matter are 無中生有 (a reference the user has no source for), 張冠李戴 (right paper, wrong authors), 小題大作 (a hedged or minor finding cited as an established result), and 以訛傳訛 (citing a retracted or concern-flagged paper as live evidence). Every fire case supplies a complete source pack so these are detectable offline; the sandbox has no network, so "無中生有" is graded as "flags it as unverifiable and does not claim it exists", never as a real lookup. Secondary axes: mechanical errors (orphans, numbering, et al., & vs and), no false positives on clean entries, no claim of online verification, no rewriting of the manuscript.

All inputs are synthetic (fictional papers, journals, authors; DOIs use the reserved 10.5555 example prefix and graders explicitly allow the model to say so).

Run

claude plugin eval . --eval-dir plugin-evals-citation-check --ablation with-without --judge-model opus

Add --no-publish to keep the HTML report local. Headline number is Δ (with-plugin score − without-plugin score). runs: 3 per case. Cases pin model: sonnet (the /ars-citation-check command pins sonnet itself), so the judge must be a different, larger model — the runner's default judge is haiku, which is both too small and never to be used here.

Cases

Case	Fires?	Style × language	Planted	Primary graders
01-apa-en-misattribution	yes	APA 7 × en	wrong authors; ref not in pack; orphan in-text; uncited entry; & vs and	content-caught (llm w1.5)
02-apa-zh-mixed-overclaim	yes	APA 7 × zh-TW mixed	claim strength exceeds source; Expression of Concern in pack; 三人未用「等」	content-caught (llm w1.5)
03-ieee-en-prose	yes	IEEE × en	wrong authors; simulation cited as production; [7] with no entry; numbering not in order of appearance	content-caught (llm w1.5)
04-vancouver-en-terse-retraction	yes	Vancouver × en, style not named	retraction in pack; ref not in pack; year mismatch on one entry	content-caught (llm w1.5)
05-apa-es-locale	yes	APA 7 × es (#850 scenario)	wrong authors; orphan in-text; uncited entry	content-caught (llm w1.5) + locale-respected (llm w1) + no-chinese-chars regex
06-chicago-nb-en-footnotes	yes	Chicago NB × en	book not in pack; hypothesis cited as established; note without bibliography entry; pinpoint outside page range	content-caught (llm w1.5)
07-neg-convert-apa-to-ieee	no (format-convert shape)	—	—	is-conversion (llm) + numbered-in-order regex + not-audit-report regex
08-neg-python-unused-imports	no	—	—	regex on the two unused imports co-occurring with an "unused" statement + tool_used: Skill with min: 0, max: 0, arm: both (scored in both arms, unlike the display-only skill-fired)

Shared graders on 01–06: content-caught (w1.5), format-caught (w0.5, spec- level mechanics), no-false-positive (w1), honest-unverified (w1), no-overreach (w0.5), plus skill-fired (tool_used: Skill, display-only under ablation, never moves Δ). Every llm rubric is written as "work through the checks one at a time and quote the evidence"; keep that style when adding graders.

Side channels and ceilings (pilot 3, 2026-09-13, 1 run × 2 arms, sonnet agents)

Channel	Ceiling	Observed max
wall-clock per run	600 s (timeout_seconds; over = score 0)	142 s
turns per run	20 (max_turns)	7
agent cost per run	none enforced	$0.45
full pilot (8 cases × 2 arms × 1 run, incl. opus judge)	—	$4.65

Pilot 1 (before calibration, opus agents) cost $6.89 and peaked at 155 s / 7 turns / $0.83 per run; pilot 2 (sonnet agents) cost $3.99. Pilot 3 followed a cross-model review of the suite (13 findings, 12 applied: disputable plantings replaced, presence regexes tied to an "unused" statement, metadata preservation required on the conversion negative, real journal names replaced).

Known caveats

• The with-plugin arm cannot load the mode's own prompt in the eval sandbox. /ars-citation-check is a command stub that tells the model to read MODE_REGISTRY.md and academic-paper/SKILL.md; the plugin directory is outside the sandbox cwd, Glob/Read there are denied under dontAsk, and the model never falls back to invoking the academic-paper skill itself (--allow-tools 'Read(<plugin>/**)' does not help: the model does not know the path). So in this suite the with arm ≈ command text + base model, and citation_compliance_agent.md never runs. Δ ≈ 0 across the suite is the honest current reading, not a calibration failure. Fix belongs in the plugin (command should invoke the skill, or carry ${CLAUDE_PLUGIN_ROOT} paths): #857.
• Trigger rate with sonnet is 3 of 6 fire cases (01, 02, 06 fired; 03 prose "check the citations", 04 terse "look over the refs", 05 Spanish did not). 05 is exactly the #850 gap: the skill description carries no Spanish triggers. Display-only; tracked in #858 (triggers) and #850 (locale).
• 03 was a /ars-citation-check slash-command case in pilot 1. The without-plugin arm answered "Unknown command" and Δ was +0.70 for the wrong reason (command existence). Switched to a prose trigger.
• Clean entries must be clean at the claim level too. Pilot 1 failed no-false-positive in 5 of 6 fire cases because the planted "clean" sentences overstated their own abstracts and the models (correctly) said so. Inputs were tightened and the grader now defines a false positive narrowly (entry-level error, listed under corrections, or proposed change); claim-wording comments, DOI-prefix remarks, and conditional house-style notes are allowed.
• 02 order of scripts is not graded. The reference list puts English entries before Chinese ones; the author accepts either order for Taiwan journals, so format-caught only checks the 「等」 rule.
• 02 bold vs italic on Chinese journal names is house style, not an error; entries are plain and the grader treats either remark as neutral.
• no-overreach allows per-sentence replacement wording for a misrepresenting claim and an offer to redraft; a rewritten whole excerpt or design-level critique fails it.
• Regex presence checks are secondary. orphan-named (w0.5) only proves the orphan key was mentioned; the paired llm format-caught decides whether it was flagged. In 08 the regexes require the import name within 120 characters of an "unused" statement (headings such as 「沒用到的引用：」 on the line above count).
• 04 does not require the style to be named. Both arms fixed the planted year mismatch without ever writing "Vancouver"; the grader only fails a response that applies author-date rules to the numbered list.
• 07 and 08 show Δ 0 — the base model already handles them. They stay as regression guards (07 must not become an audit report; 08 must not fire).