psyclaw 0.27.22 → 0.28.0

package/vendor/ars/academic-pipeline/references/pipeline_state_machine.md

@@ -0,0 +1,405 @@
+# Pipeline State Machine v2.0 — Complete Definition
+
+This document defines all legal states, transition conditions, transition actions, and exception handling for academic-pipeline v2.0.
+
+---
+
+## State Definitions
+
+### Stage States
+
+| State | Description |
+|-------|------------|
+| `pending` | Not yet started, waiting for prerequisite stage to complete |
+| `in_progress` | Currently executing |
+| `completed` | Completed, deliverables recorded |
+| `skipped` | User chose to skip (only for non-mandatory stages) |
+| `blocked` | Preconditions not met (e.g., integrity check FAIL) |
+
+### Pipeline Global States
+
+| State | Description |
+|-------|------------|
+| `initializing` | Detecting entry point and materials |
+| `running` | Pipeline executing (at least one stage is in_progress) |
+| `awaiting_confirmation` | Stage complete, waiting for user to confirm checkpoint |
+| `paused` | User paused, can resume at any time |
+| `completed` | All required stages complete, final paper produced |
+| `aborted` | User abandoned (e.g., chose to abandon after Reject) |
+
+---
+
+## State Transition Diagram (ASCII)
+
+```
+                        +-------------+
+                        | INITIALIZING|
+                        +------+------+
+                               |
+                    [Detect entry point & materials]
+                               |
+         +----------+----------+----------+----------+
+         |          |          |          |          |
+         v          v          v          v          v
+    +--------+ +--------+ +--------+ +--------+ +--------+
+    |Stage 1 | |Stage 2 | |Stg 2.5 | |Stage 3 | |Stage 4 |
+    |RESEARCH| | WRITE  | |INTEGRIT| | REVIEW | | REVISE |
+    +---+----+ +---+----+ +---+----+ +---+----+ +---+----+
+        |          |          |          |          |
+   [checkpoint]   [checkpoint]   |     [checkpoint]  |
+        |          |          |          |          |
+        v          v          v          v          v
+   +--------+ +--------+ +---+----+    |          |
+   |Stage 2 | |Stg 2.5 | |PASS?   |    |          |
+   | WRITE  | |INTEGRIT| +---+----+    |          |
+   +---+----+ +---+----+     |         |          |
+                         +----+----+    |          |
+                         |         |    |          |
+                        Yes       No    |          |
+                         |     [Fix]    |          |
+                         |   [Re-verify]|          |
+                    [checkpoint]   |    |          |
+                         |         |    |          |
+                         v         |    |          |
+                    +--------+     |    |          |
+                    |Stage 3 | <---+    |          |
+                    | REVIEW |          |          |
+                    +---+----+          |          |
+                        |               |          |
+                   [DECISION]           |          |
+                        |               |          |
+              +---------+---------+     |          |
+              |         |         |     |          |
+            Accept    Minor     Major   |          |
+              |       Revision  Revision|          |
+              |         |         |     |          |
+              |    [checkpoint]  [checkpoint]      |
+              |         |         |     |          |
+              |         v         v     |          |
+              |    +--------+ +--------+|          |
+              |    |Stage 4 | |Stage 4 ||          |
+              |    | REVISE | | REVISE ||          |
+              |    +---+----+ +---+----+|          |
+              |        |          |     |          |
+              |   [checkpoint]   [checkpoint]      |
+              |        |          |     |          |
+              |        v          v     |          |
+              |    +--------+ +--------+           |
+              |    |Stg 3'  | |Stg 3'  |           |
+              |    |RE-REV. | |RE-REV. |           |
+              |    +---+----+ +---+----+           |
+              |        |          |                 |
+              |   [DECISION]  [DECISION]            |
+              |        |          |                 |
+              |     Accept      Major               |
+              |     /Minor        |                 |
+              |        |     [checkpoint]           |
+              |        |          |                 |
+              |        |          v                 |
+              |        |     +--------+             |
+              |        |     |Stg 4'  |             |
+              |        |     |RE-REVIS|             |
+              |        |     +---+----+             |
+              |        |          |                 |
+              |   [checkpoint]  [checkpoint]        |
+              |        |          |                 |
+              v        v          v                 |
+         +----+--------+----------+-----+           |
+         |     Stage 4.5                |           |
+         |   FINAL INTEGRITY            |           |
+         +----------+------------------+           |
+                    |                               |
+               [PASS? Zero issues]                  |
+                    |                               |
+              +-----+-----+                         |
+              |           |                         |
+             Yes         No                         |
+              |        [Fix]                         |
+              |      [Re-verify]                     |
+         [checkpoint]     |                         |
+              |           |                         |
+              v           |                         |
+         +--------+       |                         |
+         |Stage 5 | <-----+                         |
+         |FINALIZE|                                 |
+         +---+----+                                 |
+             |                                      |
+        [checkpoint]---[decline Stage 6]---+        |
+             |                             |        |
+             v                             |        |
+         +--------+                        |        |
+         |Stage 6 |                        |        |
+         |PROCESS |                        |        |
+         |SUMMARY |                        |        |
+         +---+----+                        |        |
+             |                             |        |
+   [terminal acknowledgement]              |        |
+             |                             |        |
+             v                             |        |
+        +---------+                        |        |
+        |COMPLETED| <----------------------+        |
+        +---------+                                 |
+```
+
+---
+
+## Legal State Transitions
+
+### Normal Flow Transitions
+
+| From | To | Precondition | Action |
+|------|----|-------------|--------|
+| INIT | Stage 1 | User confirms starting from Stage 1 | Detect mode preference, launch deep-research |
+| INIT | Stage 2 | User has research materials, confirms skipping Stage 1 | Detect materials, launch academic-paper |
+| INIT | Stage 2.5 | User has complete paper | Launch integrity_verification_agent |
+| INIT | Stage 3 | User has verified paper + integrity report | Confirm paper language/domain, launch reviewer |
+| INIT | Stage 4 | User has review comments | Confirm paper + review comments, launch revision |
+| INIT | Stage 5 | User has final draft for format conversion | Confirm format requirements, launch format-convert |
+| Stage 1 | **checkpoint** | Stage 1 completed | Wait for user confirmation |
+| checkpoint | Stage 2 | User confirms | handoff RQ Brief + Methodology Blueprint + Bibliography + Synthesis |
+| Stage 2 | **checkpoint** | Stage 2 completed, Paper Draft produced | Wait for user confirmation |
+| checkpoint | Stage 2.5 | User confirms | Pass Paper Draft to integrity agent |
+| Stage 2.5 | **checkpoint** | PASS, or recorded Integrity Check FAIL Loop resolution (§ below) | Wait for user confirmation |
+| Stage 2.5 | Stage 2.5 (retry) | FAIL | Fix issues, re-verify (max 3 rounds) |
+| checkpoint | Stage 3 | User confirms | Pass verified paper to reviewer |
+| Stage 3 | **checkpoint** | Decision produced | Wait for user confirmation |
+| checkpoint | Stage 4 | Decision = Minor/Major, user confirms | Pass Revision Roadmap |
+| checkpoint | Stage 4.5 | Decision = Accept, user confirms | Skip revision, go directly to final verification |
+| Stage 4 | **checkpoint** | Stage 4 completed | Wait for user confirmation |
+| checkpoint | Stage 3' | User confirms | Pass Revised Draft + Original (pre-revision) Draft (#576 §3.1 Phase 2A comparison base) + Response to Reviewers + Editorial Decision Letter (#539 Judge Record input) + Round-1 review findings (Schema 6 reports — #576 §4 level-3 layer) + Round-1 Revision Roadmap + apply report(s) with their paired revision patch/diff files (#390/#576 §11 — the two travel together) + Round-1 Reviewer Configuration Cards (yardstick continuity). Re-review-mode transfer (default); a user-requested fresh full review at 3' passes Revised Draft + available context only (no Roadmap/cards; full mode runs field_analyst) |
+| Stage 3' | **checkpoint** | Decision produced | Wait for user confirmation |
+| checkpoint | Stage 4.5 | Decision = Accept/Minor, user confirms | Pass final draft to final verification |
+| checkpoint | Stage 4' | Decision = Major, user confirms | Pass new Revision Roadmap |
+| Stage 4' | **checkpoint** | Stage 4' completed | Wait for user confirmation |
+| checkpoint | Stage 4.5 | User confirms | Pass revised draft to final verification |
+| Stage 4.5 | **checkpoint** | PASS (zero issues), or recorded Integrity Check FAIL Loop resolution (§ below) | Wait for user confirmation |
+| Stage 4.5 | Stage 4.5 (retry) | FAIL | Fix issues, re-verify (max 3 rounds) |
+| checkpoint | Stage 5 | User confirms (MANDATORY — the Stage 5 entry gate; see § Stage 5 boundary semantics) | Pass final accepted draft; record the finalization-format decision (citation style) |
+| Stage 5 | **checkpoint** | Stage 5 completed, Final Paper delivered | Wait for user confirmation (FULL — never SLIM; see § Stage 5 boundary semantics) |
+| checkpoint | Stage 6 | User confirms | Dispatch Process Summary per `process_summary_protocol.md` |
+| checkpoint | completed | User declines Stage 6 | Mark Stage 6 `skipped` (non-mandatory stage); set pipeline global state `completed` |
+| Stage 6 | **terminal checkpoint** | Process Record delivered | Wait for terminal acknowledgement (see § Stage 6 terminal semantics) |
+| terminal checkpoint | completed | User acknowledges (`finish` / `end` / `done` / `confirm`, or an unambiguous natural-language equivalent) | Mark Stage 6 `completed`; set pipeline global state `completed` |
+
+### Special Flow Transitions
+
+| From | To | Precondition | Action |
+|------|----|-------------|--------|
+| Stage 3 (Reject) | Stage 2 | User chooses to restructure | Clear Stage 2-3 state, preserve Stage 1 materials, restart Stage 2 |
+| Stage 3 (Reject) | ABORT | User chooses to abandon | Save all produced materials, mark pipeline aborted |
+| Stage 3' (Major) | Stage 4' | User confirms | Last revision opportunity |
+| Stage 4' | Stage 4.5 | Revision complete | Go directly to final verification (no return to review) |
+| Any stage | PAUSED | User says "pause" or "stop here" | Save pipeline state |
+| PAUSED | Previous stage | User returns to continue | Restore pipeline state, display Dashboard |
+
+### Prohibited Transitions (Illegal)
+
+| From | To | Reason |
+|------|----|--------|
+| Stage 1 | Stage 3 | Cannot skip Stage 2 and 2.5 (unless mid-entry + has paper) |
+| Stage 2 | Stage 3 | **Cannot skip Stage 2.5 (integrity check is mandatory)** |
+| Stage 4 | Stage 5 | Cannot skip RE-REVIEW (revision must be re-reviewed) |
+| Stage 3' | Stage 5 | **Cannot skip Stage 4.5 (final integrity check is mandatory)** |
+| Stage 4' | Stage 3' | Cannot return to RE-REVIEW (max 1 round of RE-REVISE) |
+| Stage 5 | Stage 3 | Cannot roll back (no review after FINALIZE) |
+| completed | in_progress | Completed stages cannot restart |
+
+---
+
+## Stage 5 and Stage 6 Boundary Semantics (#528)
+
+The two boundaries below were under-specified before v3.17 (different runtimes could resolve them differently). This section is the authority; `SKILL.md` and `pipeline_orchestrator_agent.md` mirror it.
+
+### Stage 5 boundary semantics
+
+"Before finalization (Stage 5): always MANDATORY" refers to exactly ONE checkpoint: the **Stage 5 entry gate** — the checkpoint between Stage 4.5 PASS and the Stage 5 dispatch. It is MANDATORY because it carries the finalization decisions:
+
+- explicit confirmation to proceed to finalization (no auto-advance);
+- the finalization-format decision: citation style (APA 7.0 / Chicago / IEEE, ...) — the "Stage 5 finalization format" pending decision the passport-reset machinery records at this boundary.
+
+Transition state: `awaiting_confirmation` → on user confirmation → Stage 5 `in_progress`.
+
+#660 and #672 surface inside this same one checkpoint, after the same exact
+Stage 4.5 PASS and before the confirmation. The orchestrator runs #660 first and
+#672 second against the identical accepted-draft artifact ID/SHA-256. They do
+not create a second checkpoint or transition state. A schema-valid #660 degraded
+artifact (exit 1) is preserved; a #672 failure writes no carrier and records only
+bounded `ADVISORY_UNAVAILABLE:<CODE>`. Neither result blocks/delays confirmation,
+changes Stage 4.5, adds remediation routing, or changes Stage-5 dispatch.
+
+Any manuscript revision stales both carriers and returns through integrity before
+the fixed #660-then-#672 sequence reruns on the new accepted bytes. Reusing only
+one old carrier is invalid handoff cargo, not a new state-machine branch.
+
+Other confirmations near Stage 5 are NOT this MANDATORY boundary:
+
+1. The in-stage interactions of the Stage 5 output process — the "Need LaTeX?" question (Step 3) and the content confirmation before the final PDF (Step 4) — are part of Stage 5 execution, not pipeline checkpoints; they are asked during the stage, never at the gate.
+2. The **Stage 5 completion checkpoint** (Final Paper delivered, before Stage 6) follows the global stage-completion rule: it is a FULL checkpoint — never SLIM, because final-deliverable acceptance must not be downgraded — but it is not on the MANDATORY list.
+
+### Stage 6 terminal semantics
+
+Stage 6 is a non-mandatory stage (it is absent from the orchestrator's non-skippable list). At the Stage 5 completion checkpoint the user may decline it: Stage 6 is marked `skipped` and the pipeline still terminates `completed` (the Final Paper was already produced at Stage 5).
+
+When Stage 6 runs, its completion is the pipeline's **terminal checkpoint**:
+
+1. After delivering the Process Record (MD + PDF per the user's language choice), the orchestrator prompts for a terminal acknowledgement.
+2. Terminal acknowledgement vocabulary: `finish` / `end` / `done` / `confirm`, or an unambiguous natural-language equivalent that accepts the deliverables. Change requests (the other language version, content corrections) keep Stage 6 `in_progress` — they are not acknowledgements.
+3. On acknowledgement: state_tracker marks Stage 6 `completed` and sets the pipeline global state to `completed`. This is the terminal transition — there is no next stage.
+4. After `completed`, no stage transition is legal (see Prohibited Transitions). New requests start a new pipeline run or a targeted single-skill invocation (mid-entry).
+
+### Post-terminal adjudication-activity side channel (#673)
+
+The ordinary state machine is authoritative and always terminates first. A
+`completed` run durably records Stage 6 with status `completed` or `skipped` as
+defined above. An `aborted` run durably records its first replayable terminal
+stage. The closed replayable activity vocabulary is Stage 1, 2, 2.5, 3,
+3-prime, 4, 4-prime, 4.5, 5, and 6; there is no Stage 0.
+
+Only after that terminal write succeeds, and only for an explicitly selected
+local store, may the orchestrator invoke the state tracker's #673
+post-terminal sequence. It passes explicit state/artifact-root paths plus the
+explicit five-row `pending_adjudication_activity_bindings[]` value to
+`seal_terminal_inventory(state_path, artifact_root, pending_bindings)`. Pending
+captured artifact bindings carry id, role, group id,
+`artifact_group_stage`, and relative path but no hash; the helper computes raw
+hashes. Non-applicable/unavailable rows carry empty artifacts and a closed
+reason. The helper neither reads the pending field itself nor infers or scans
+for sources.
+
+The helper seals the exact root `adjudication_activity_sources` inventory while
+leaving terminal state, stage, and status byte-semantically unchanged. The
+terminal file's root `run_id` plus that sealed root inventory are exact
+source/run authority; pending rows are not. `build-input` can project only the
+sealed inventory and cannot accept caller-reported hashes. Idempotent append
+and optional render follow. Any post-terminal failure is advisory, creates no
+substitute record, and cannot change the already-durable terminal state.
+
+This side channel is never a legal transition, precondition, gate, verdict,
+checkpoint input, passport/handoff/Process Record field, or model/observer/
+compliance input. It uses no live model, judge, eval, network/API, ambient
+clock, directory scan, or glob. Producer details are single-homed in
+`../agents/state_tracker_agent.md` § "Adjudication-activity metadata".
+
+---
+
+## Material Dependency Matrix
+
+| Material | Produced At | Consumed At | Required/Recommended |
+|----------|-----------|-------------|---------------------|
+| RQ Brief | Stage 1 | Stage 2 (Phase 0) | Recommended |
+| Methodology Blueprint | Stage 1 | Stage 2 (Phase 0) | Recommended |
+| Bibliography | Stage 1 | Stage 2 (Phase 1) | Recommended |
+| Synthesis Report | Stage 1 | Stage 2 (Phase 3) | Recommended |
+| `preregistration-artifact/1.0` sidecar + provided named companion | Stage 1 shell-capable dispatch | Academic-paper intake, every handoff, Stage 4.5/#672 checkpoint | **Receipt required** (content may be unavailable); validate and carry exact bytes, never infer/rebuild/template-substitute |
+| Paper Draft | Stage 2 | Stage 2.5 (input) | **Required** |
+| **Integrity Report (Pre)** | **Stage 2.5** | **Stage 3 (prerequisite)** | **Required** |
+| **Verified Paper Draft** | **Stage 2.5** | **Stage 3 (Phase 0) + Stage 3' (re-review mode — the original pre-revision manuscript, #576 1.1 §3.1 Phase 2A comparison base)** | **Required, including Stage 3' consumption; current manifest and bundle fail closed if absent** |
+| Review Reports (x5) | Stage 3 | Stage 4 (input) + Stage 3' (re-review mode — the #576 §4 level-3 driving-finding criterion layer; absent → transported Schema 7 fields alone, `[ROUND1-FINDINGS-ABSENT]`) | Required (Stage 3' consumption Recommended — §11 presence policy) |
+| Editorial Decision | Stage 3 | Stage 4 (input) | Required |
+| Revision Roadmap | Stage 3 | Stage 4 (input) + Stage 3' (re-review mode — verification checklist basis; the fresh-full-review branch consumes none) | Required (Stage 3' consumption re-review-mode-only) |
+| Revised Draft | Stage 4 | Stage 3' (Phase 0) | Required |
+| Response to Reviewers | Stage 4 | Stage 3' (input) | Recommended |
+| Editorial Decision Letter (Schema 6) | Stage 3 | Stage 3' (input) | Recommended (#539 — its Review Panel Provenance block feeds the Judge Record; absent → "unknown (provenance block absent)") |
+| Apply report(s) + paired revision patch/diff files (#390 sidecar + patch JSON) | Stage 4 | Stage 3' (input) + Stage 4.5 (input) | Recommended (patch-apply rounds only; the two travel TOGETHER — reports without their paired patches is `manifest_incomplete` at Stage 3' dispatch; chain verified by the #576 §11 ordered-chain rule — first report's `base_draft_hash` = original draft, each subsequent = predecessor's `output_draft_hash`, last = Revised Draft; see `re_review_mode_protocol.md` § Input Manifest) |
+| Traceability sidecar (`shared/contracts/re_review/traceability.schema.json`) | Stage 3' | Stage 4.5 (input — frozen `previously_missed`/`indeterminate` records, #576 §8; direct on Accept/Minor, through 4' on Major) | Recommended (contract-mode Stage 3' only; a `[LEGACY-NO-CONTRACT]` run produces none — visible degradation at 4.5, not a block) |
+| Reviewer Configuration Cards | Stage 3 (field_analyst) | Stage 3' (re-review mode — yardstick continuity; field_analyst NOT re-run) | Recommended (absent → visible regeneration fallback, `re_review_mode_protocol.md` § Yardstick Continuity; a fresh full review at 3' regenerates by definition) |
+| **Re-Review Report** | **Stage 3'** | **Stage 4' (input)** | **Required (if Major)** |
+| **Re-Revised Draft** | **Stage 4'** | **Stage 4.5 (input)** | **Required (if executed)** |
+| **Integrity Report (Final)** | **Stage 4.5** | **Stage 5 (prerequisite)** | **Required** |
+| Final Paper | Stage 5 | User (delivery) | Required |
+| Process Record | Stage 6 | User (delivery) | Optional (Stage 6 is skippable) |
+
+---
+
+## Exception State Handling
+
+### Timeout
+
+If a stage shows no progress for an extended period (e.g., Socratic mode exceeds 15 rounds without convergence):
+1. state_tracker marks the stage as `stalled`
+2. orchestrator provides options:
+   - Switch mode (socratic -> full)
+   - Narrow scope
+   - Skip this stage (non-mandatory stages only)
+
+### Missing Materials
+
+If required materials are found missing during transition:
+1. state_tracker reports the material gap
+2. orchestrator suggests returning to the stage that produces that material
+3. User can choose: backfill / skip (at own risk, but cannot skip integrity checks)
+
+### Integrity Check FAIL Loop
+
+If Stage 2.5 or 4.5 corrections exceed 3 rounds without passing:
+1. List all unverifiable items
+2. User decides:
+   - Manually handle unverifiable items
+   - Remove unverifiable citations
+   - Continue to next stage (with "partially unverified" warning)
+
+### Session Interruption
+
+If the user leaves and returns:
+1. orchestrator displays Progress Dashboard
+2. Confirm whether to continue from breakpoint
+3. Check if any outdated materials need refreshing
+
+---
+
+## Revision Loop Rules (v2.0)
+
+### Simplified Revision Cycle
+
+```
+v2.0's revision cycle is simpler and more explicit than v1.0:
+
+Stage 3 (First REVIEW)
+  -> Decision: Accept -> Stage 4.5
+  -> Decision: Minor/Major -> Stage 4
+      -> Stage 4 (REVISE)
+          -> Stage 3' (RE-REVIEW, verification)
+              -> Decision: Accept/Minor -> Stage 4.5
+              -> Decision: Major -> Stage 4' (last revision)
+                  -> Stage 4.5 (go directly to final verification, no return to review)
+
+Maximum 1 round of RE-REVISE, no infinite loops.
+Unresolved issues -> Acknowledged Limitations.
+```
+
+### Differences from v1.0
+
+| v1.0 | v2.0 |
+|------|------|
+| Max 2 review-revise cycles | Fixed 2 reviews (Stage 3 + Stage 3') + max 1 RE-REVISE |
+| No integrity check | Mandatory Pre-review + Final integrity check |
+| 4 reviewers | 5 reviewers (+Devil's Advocate) |
+| Can skip any stage | Stage 2.5 and 4.5 cannot be skipped |
+| No mandatory checkpoints | Every stage requires a checkpoint |
+
+## Reset-boundary transitions (v3.6.3, flag-gated)
+
+When `ARS_PASSPORT_RESET=1`, every FULL checkpoint carries an implicit state transition to a terminal `awaiting_resume` state. The next stage only starts when a new session posts `resume_from_passport=<hash>`.
+
+Transition semantics:
+
+```
+Stage N [working]
+  -> FULL checkpoint
+    -> [flag OFF]  Stage N+1 [working]           (pre-v3.6.3 continuation)
+    -> [flag ON]   append boundary entry -> awaiting_resume
+         -> resume_from_passport=<hash>
+              -> append resume entry (consumes_hash=<hash>)
+              -> Stage N+1 [working]              (fresh session, passport-loaded)
+```
+
+Iron rules:
+
+- `awaiting_resume` is not persisted in `state_tracker`; it is computed from the passport ledger. A `boundary` entry with hash `H` is awaiting resume iff no later `resume` entry in `reset_boundary[]` carries `consumes_hash == H`. Single pass over the ledger, no out-of-band state.
+- `systematic-review` under flag ON cannot transition `Stage N → Stage N+1` without a fresh-session resume. In-session continuation is refused.
+- Other modes under flag ON allow in-session continuation as a fallback, but the orchestrator must still load Stage N+1 input strictly from the passport (no replay of prior turns).
+- SLIM checkpoints never enter `awaiting_resume`.
+- MANDATORY checkpoints enter `awaiting_resume` when they are also FULL and flag is ON. Integrity gates remain MANDATORY; the reset does not downgrade them. The `### Resume Instruction` subsection emitted alongside `[PASSPORT-RESET: ...]` carries the passport file path and resume command — it does NOT carry the user decision prompt. The decision prompt happens on resume, after the fresh session loads the passport (see next rule).
+- If a `boundary` entry carries `pending_decision`, `next` is advisory only. The user's branch choice happens AFTER `resume_from_passport=<hash>` in the fresh session, never in the reset checkpoint itself. The orchestrator re-prompts the user in the new session before transitioning to any `Stage N+1`. The `resume` entry records the chosen branch via `chosen_branch`. Actual routing comes from the matched option's `next_stage`/`next_mode`; `next` is a fallback default only.
+
+See [`passport_as_reset_boundary.md`](passport_as_reset_boundary.md) for the full protocol.

package/vendor/ars/academic-pipeline/references/plagiarism_detection_protocol.md

@@ -0,0 +1,239 @@
+# Plagiarism Detection Protocol — Phase D Originality Verification Protocol
+
+This document defines the complete execution protocol for `integrity_verification_agent`'s Phase D (originality verification), including paragraph-level comparison, self-plagiarism check, AI-generated text characteristic detection, severity grading, and tool limitation disclaimers.
+
+---
+
+## Phase D Overview: Originality Verification
+
+Phase D's purpose is to perform originality screening on body text content before paper submission for review and after revision completion. Unlike Phases A-C which focus on "whether citations and data are correct," Phase D focuses on "whether the body text itself is originally written."
+
+**Core principle: Heuristic screening, not final determination.** This protocol uses WebSearch for publicly available literature comparison. Results are preliminary screening signals and do not equate to conclusions from professional plagiarism detection software.
+
+---
+
+## D1: Paragraph-Level Originality Check
+
+### D1.1 Characteristic Sentence Extraction
+
+```
+For each paragraph in the paper body text:
+1. Identify the paragraph's topic and core argument
+2. Extract 1-2 "characteristic sentences"
+   - Priority selection: Sentences containing specific data, proper nouns, or unique arguments
+   - Avoid: Generic academic boilerplate (e.g., "This study aims to...")
+3. Record the characteristic sentence's paragraph location (section + paragraph number)
+```
+
+### D1.2 WebSearch Comparison
+
+```
+For each extracted characteristic sentence:
+1. Use the characteristic sentence (or key fragment) as a WebSearch query
+   - Search term: Enclose 8-12 consecutive words in quotation marks
+   - Supplementary search: Remove quotes to check for paraphrased versions
+2. Review the top 5-10 search results
+3. Compare text similarity between original and search results
+```
+
+### D1.3 Comparison Result Grading
+
+| Grade | Code | Definition | Determination Criteria |
+|-------|------|-----------|----------------------|
+| Original | `ORIGINAL` | No similar expression found in public literature | WebSearch returns no related matches |
+| Common Knowledge | `COMMON_KNOWLEDGE` | The knowledge is a widely accepted fact in the field | Multiple sources express the same fact in different ways |
+| Paraphrase | `PARAPHRASE` | Expresses same viewpoint as a source but with clearly different wording | Semantically similar but significantly different sentence structure and word choice, with citation |
+| Close Match | `CLOSE_MATCH` | Highly similar wording to a source, with only a few words substituted | Nearly identical sentence structure, with only synonym substitutions or word order changes |
+| Verbatim | `VERBATIM` | Identical or nearly identical text to a source | 20+ consecutive identical words without quotation marks |
+
+### D1.4 Sampling Rate Requirements
+
+| Operating Mode | Minimum Sampling Rate | Description |
+|---------------|----------------------|------------|
+| Mode 1 (pre-review) | **30%** | At least 30% of all body text paragraphs checked |
+| Mode 2 (final-check) | **50%** | At least 50% of all body text paragraphs checked |
+
+**Sampling strategy**:
+- Priority check: Literature Review, Background, Discussion and other high-risk sections
+- Must cover: At least 1 paragraph sampled from each major chapter
+- Random supplement: Beyond priority paragraphs, randomly sample paragraphs to reach minimum sampling rate
+- Revised paragraphs: In Mode 2, all paragraphs newly added or substantially modified during revision must be checked 100%
+
+---
+
+## D2: Self-Plagiarism Check
+
+### D2.1 Author's Existing Publications Search
+
+```
+Prerequisite: User provides author name(s)
+
+For each author (or primary author):
+1. WebSearch: "author name" + research area keywords
+2. Identify author's existing publication list (Google Scholar profile preferred)
+3. Record existing publications related to the current paper's topic
+```
+
+### D2.2 Comparison Items
+
+```
+Compare current paper with author's existing publications (focus on these areas):
+1. Methodology descriptions:
+   - Is the research design description verbatim identical to prior work?
+   - Are data collection and analysis method descriptions directly copied?
+2. Results narratives:
+   - Are textual descriptions of results reused?
+   - Are table/figure description texts identical?
+3. Theoretical framework:
+   - Are literature review paragraphs transferred wholesale?
+```
+
+### D2.3 Legitimate Self-Citation vs. Self-Plagiarism Determination Criteria
+
+| Scenario | Determination | Description |
+|----------|--------------|------------|
+| Cites prior work and restates in new language | **Legitimate self-citation** | Normal academic practice — has citation and paraphrasing |
+| Cites prior work but verbatim copies original text | **Self-plagiarism** | Even with citation, extensive verbatim copying is unacceptable |
+| Content highly similar to prior work without citing it | **Self-plagiarism** | Conceals relationship with prior work |
+| Uses prior work's data but re-analyzes | **Legitimate** | Secondary analysis is a legitimate research method — must clearly state this |
+| Methodology reuses prior work's standardized description | **Gray area** | Standardized experimental procedure descriptions allow high similarity, but citing prior work is recommended |
+
+---
+
+## D3: AI-Generated Text Characteristic Detection
+
+**Important disclaimer: This section is a checklist, not a determination tool.** AI text detection technology is not yet mature, and any judgment based on text characteristics has a high false-positive risk. The following indicators are for reference only and should not serve as the basis for final determination.
+
+### D3.1 Typical AI Writing Pattern Indicators
+
+| # | Indicator | Description | Observation Method |
+|---|-----------|------------|-------------------|
+| 1 | Excessive smoothness | Abnormally uniform sentence fluency throughout, lacking natural writing rhythm variation | Compare whether writing style across chapters is overly consistent |
+| 2 | Lack of specificity | Arguments remain at conceptual level, lacking specific numbers, cases, or personal research experience | Check for "for example" followed by vague content |
+| 3 | Formulaic transitions | Heavy use of "Furthermore," "Moreover," "It is worth noting that" and similar transitions | Count the variety and frequency of transition phrases |
+| 4 | Excessive parallelism | Highly symmetric paragraph structures (e.g., every paragraph follows: claim -> evidence -> summary) | Observe whether paragraph structure mechanically repeats |
+| 5 | Hedging overload | Excessive use of "may," "could," "might," "it is possible that" to avoid definitive positions | Check whether author over-hedges even on their own research results |
+| 6 | Citation-argument gap | Literature is cited but the cited content is not organically integrated with the author's arguments | Remove citations — does the paragraph's argument still hold? |
+
+### D3.2 Handling Approach
+
+```
+If the paper triggers 2 or more AI writing indicators:
+1. Flag in the verification report as "AI writing characteristic alert"
+2. List the specific indicators triggered and corresponding paragraphs
+3. Do NOT make a "whether it is AI-generated" determination
+4. Recommend the user review the flagged paragraphs and consider adjusting writing style
+```
+
+---
+
+## Severity Grading
+
+| Level | Code | Definition | Trigger Conditions |
+|-------|------|-----------|-------------------|
+| **Critical** | `CRITICAL` | Severe academic misconduct, sufficient for retraction | Verbatim plagiarism (>20 consecutive identical words without citation); fabricated citations (citing nonexistent sources to support plagiarized content) |
+| **Serious** | `SERIOUS` | Significant originality problems, requiring major revisions | Multiple close paraphrases without citing sources; extensive undisclosed self-plagiarism |
+| **Moderate** | `MODERATE` | Individual paragraphs need rewriting | Individual paragraphs inadequately paraphrased (1-2 instances of `CLOSE_MATCH`); methodology description overly similar to prior work |
+| **Minor** | `MINOR` | Does not affect academic integrity but improvement recommended | Excessive generic academic boilerplate; AI writing characteristic alerts (informational only, does not affect verdict) |
+
+### Severity-to-Verdict Mapping
+
+| Severity | Impact on Verdict |
+|---------|-----------------|
+| CRITICAL | Immediate FAIL, listed as highest priority correction item |
+| SERIOUS | FAIL, must fix and re-verify |
+| MODERATE | FAIL, must fix |
+| MINOR | Does not affect PASS/FAIL verdict, noted in report |
+
+---
+
+## Tool Limitation Disclaimer
+
+This protocol's originality verification has the following inherent limitations that users must be aware of:
+
+| # | Limitation | Description |
+|---|-----------|------------|
+| 1 | **Not professional plagiarism detection software** | This protocol uses WebSearch for heuristic comparison, not Turnitin, iThenticate, or other professional tools — cannot calculate precise text overlap rates |
+| 2 | **Limited coverage** | Can only compare publicly searchable literature (open access, preprints, web pages) — cannot search full-text databases behind paywalls |
+| 3 | **Language limitation** | Cross-language plagiarism (e.g., plagiarism via translation) is difficult to detect |
+| 4 | **Sampling, not exhaustive** | Limited by efficiency, only 30%-50% of paragraphs are sampled — missed detection risk exists |
+| 5 | **Time sensitivity** | Search results change over time; newly published literature may not be in search scope |
+| 6 | **AI detection unreliable** | D3's AI writing indicators are heuristic alerts with high false-positive rates and should not serve as determination basis |
+
+**Recommendation**: This protocol's results serve as preliminary screening. It is recommended to use professional plagiarism detection tools (such as Turnitin / iThenticate) for complete duplicate checking before formal submission.
+
+---
+
+## Output Format Template
+
+```markdown
+## Phase D: Originality Verification Results
+
+### Verification Parameters
+- Operating mode: [Mode 1 pre-review / Mode 2 final-check]
+- Total body text paragraphs: X
+- Paragraphs sampled: Y (sampling rate: Z%)
+- Author self-plagiarism check: [Executed / Not executed (author information not provided)]
+
+### D1 Paragraph-Level Comparison Results Summary
+
+| Grade | Paragraph Count | Proportion |
+|-------|----------------|-----------|
+| ORIGINAL | X | X% |
+| COMMON_KNOWLEDGE | X | X% |
+| PARAPHRASE | X | X% |
+| CLOSE_MATCH | X | X% |
+| VERBATIM | X | X% |
+
+### D2 Self-Plagiarism Check Results
+
+| # | Current Paper Paragraph | Existing Publication | Similarity Type | Determination |
+|---|------------------------|---------------------|----------------|--------------|
+| 1 | §X.X, paragraph Y | Author (Year), Title | Methodology description similar | Legitimate self-citation / Self-plagiarism |
+
+### D3 AI Writing Characteristic Alerts
+
+| # | Indicator | Triggered Paragraph | Description |
+|---|-----------|---------------------|------------|
+| 1 | [Indicator name] | §X.X | [Specific observation] |
+
+Indicators triggered: X / 6 ([Below threshold, not flagged / Threshold reached, user review recommended])
+
+### Phase D Issue List
+
+| # | Severity | Type | Location | Issue Description | Matching Source | Recommended Action |
+|---|----------|------|----------|------------------|----------------|-------------------|
+| 1 | CRITICAL | VERBATIM | §X.X, paragraph Y | N consecutive words identical to source | [URL] | Rewrite or add quotation marks for direct quote |
+| 2 | SERIOUS | CLOSE_MATCH | §X.X, paragraph Y | Highly similar wording, only a few words substituted | [URL] | Rewrite and add citation |
+| 3 | MODERATE | Self-plagiarism | §X.X, paragraph Y | Methodology description verbatim identical to prior work | Author (Year) | Rewrite and cite prior work |
+
+### Tool Limitation Disclaimer
+
+> This originality verification uses WebSearch for heuristic comparison and is not professional plagiarism detection software (such as Turnitin / iThenticate). Coverage is limited to publicly searchable literature, with a sampling rate of [Z]%, and there is a risk of missed detection. These results serve as preliminary screening; it is recommended to use professional plagiarism detection tools for complete duplicate checking before formal submission.
+```
+
+---
+
+## Relationship with Other Phases
+
+| Phase | Focus | Relationship with Phase D |
+|-------|-------|--------------------------|
+| Phase A: Reference Verification | Whether references exist and are correct | A verifies sources, D verifies body text; if D finds VERBATIM without citation, it may also reveal A3 dangling citation issues |
+| Phase B: Citation Context Verification | Whether citations accurately reflect original text | B checks "whether cited content is correct," D checks "whether uncited content is original" |
+| Phase C: Data Verification | Whether statistical data is correct | C and D are complementary: C verifies data, D verifies text |
+
+---
+
+## Reproducibility Requirements
+
+To ensure the originality verification process is reproducible:
+
+1. **Standardized search strategy**: Use the same search template for each characteristic sentence
+   - Search term 1: `"key fragment" (8-12 words, in quotes)`
+   - Search term 2: `keyword combination (without quotes, to match paraphrases)`
+
+2. **Explicit determination criteria**: Each grade (ORIGINAL through VERBATIM) has clear determination criteria, not relying on subjective feeling
+
+3. **Complete records**: Search terms, search results, and determination rationale for each sampled paragraph are recorded in the Audit Trail
+
+4. **Timestamps**: Report includes execution time, as search results change over time