pearmut 1.0.0__tar.gz → 1.0.1__tar.gz

{pearmut-1.0.0 → pearmut-1.0.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pearmut
-Version: 1.0.0
+Version: 1.0.1
 Summary: A tool for evaluation of model outputs, primarily MT.
 Author-email: Vilém Zouhar <vilem.zouhar@gmail.com>
 License: MIT
@@ -19,17 +19,10 @@ Provides-Extra: dev
 Requires-Dist: pytest; extra == "dev"
 Dynamic: license-file
 
-# Pearmut 🍐
+# 🍐Pearmut &nbsp; &nbsp; [![PyPi version](https://badgen.net/pypi/v/pearmut/)](https://pypi.org/project/pearmut) [![PyPI download/month](https://img.shields.io/pypi/dm/pearmut.svg)](https://pypi.python.org/pypi/pearmut/) [![PyPi license](https://badgen.net/pypi/license/pearmut/)](https://pypi.org/project/pearmut/) [![build status](https://github.com/zouharvi/pearmut/actions/workflows/test.yml/badge.svg)](https://github.com/zouharvi/pearmut/actions/workflows/test.yml)
 
 **Platform for Evaluation and Reviewing of Multilingual Tasks**: Evaluate model outputs for translation and NLP tasks with support for multimodal data (text, video, audio, images) and multiple annotation protocols ([DA](https://aclanthology.org/N15-1124/), [ESA](https://aclanthology.org/2024.wmt-1.131/), [ESA<sup>AI</sup>](https://aclanthology.org/2025.naacl-long.255/), [MQM](https://doi.org/10.1162/tacl_a_00437), and more!).
 
-[![PyPi version](https://badgen.net/pypi/v/pearmut/)](https://pypi.org/project/pearmut)
-&nbsp;
-[![PyPI download/month](https://img.shields.io/pypi/dm/pearmut.svg)](https://pypi.python.org/pypi/pearmut/)
-&nbsp;
-[![PyPi license](https://badgen.net/pypi/license/pearmut/)](https://pypi.org/project/pearmut/)
-&nbsp;
-[![build status](https://github.com/zouharvi/pearmut/actions/workflows/test.yml/badge.svg)](https://github.com/zouharvi/pearmut/actions/workflows/test.yml)
 
 <img width="1000" alt="Screenshot of ESA/MQM interface" src="https://github.com/user-attachments/assets/71334238-300b-4ffc-b777-7f3c242b1630" />
 
@@ -52,6 +45,8 @@ Dynamic: license-file
 - [Terminology](#terminology)
 - [Development](#development)
 - [Citation](#citation)
+- [Changelog](#changelog)
+
 
 ## Quick Start
 
@@ -111,7 +106,9 @@ Campaigns are defined in JSON files (see [examples/](examples/)). The simplest c
 }
 ```
 
-Each item has to have `src` (string) and `tgt` (dictionary from model names to strings, even for a single model evaluation).
+Each item has to have `tgt` (dictionary from model names to strings, even for a single model evaluation).
+Optionally, you can also include `src` (source string) and/or `ref` (reference string).
+If neither `src` nor `ref` is provided, only the model outputs will be displayed.
 For full Pearmut functionality (e.g. automatic statistical analysis), add `item_id` as well.
 Any other keys that you add will simply be stored in the logs.
 
@@ -145,6 +142,40 @@ The `shuffle` parameter in campaign `info` controls this behavior:
 }
 ```
 
+### Custom Score Sliders
+
+For multi-dimensional evaluation tasks (e.g., assessing fluency on a Likert scale), you can define custom sliders with specific ranges and steps:
+
+```python
+{
+  "info": {
+    "assignment": "task-based",
+    "protocol": "ESA",
+    "sliders": [
+      {"name": "Fluency", "min": 0, "max": 5, "step": 1},
+      {"name": "Adequacy", "min": 0, "max": 100, "step": 1}
+    ]
+  },
+  "campaign_id": "my_campaign",
+  "data": [...]
+}
+```
+
+When `sliders` is specified, only the custom sliders are shown. Each slider must have `name`, `min`, `max`, and `step` properties. All sliders must be answered before proceeding.
+
+### Custom Instructions
+
+Set campaign-level instructions using the `instructions` field in `info` (supports HTML).
+Instructions default to protocol-specific ones (DA: scoring, ESA: error spans + scoring, MQM: error spans + categories + scoring).
+```python
+{
+  "info": {
+    "protocol": "DA",
+    "instructions": "Rate translation quality on a 0-100 scale.<br>Pay special attention to document-level phenomena."
+  }
+}
+```
+
 ### Pre-filled Error Spans (ESA<sup>AI</sup>)
 
 Include `error_spans` to pre-fill annotations that users can review, modify, or delete:
@@ -263,7 +294,7 @@ All items must contain outputs from all models for this assignment type to work
 **How it works:**
 1. Initial phase: Each model gets `dynamic_first` annotations with fully random contrastive evaluation
 2. Dynamic phase: After the initial phase, top `dynamic_top` models (by average score) are identified
-3. Contrastive evaluatoin: From the top N models, `dynamic_contrastive_models` models are randomly selected for each item
+3. Contrastive evaluation: From the top N models, `dynamic_contrastive_models` models are randomly selected for each item
 4. Item prioritization: Items with the least annotations for the selected models are prioritized
 5. Backoff: With probability `dynamic_backoff`, uniform random selection is used instead to maintain exploration
 
@@ -289,6 +320,7 @@ The `users` field accepts:
 }
 ```
 
+
 ### Multimodal Annotations
 
 Support for HTML-compatible elements (YouTube embeds, `<video>` tags, images). Ensure elements are pre-styled. See [examples/multimodal.json](examples/multimodal.json).
@@ -418,3 +450,61 @@ If you use this work in your paper, please cite as following.
 ```
 
 Contributions are welcome! Please reach out to [Vilém Zouhar](mailto:vilem.zouhar@gmail.com).
+
+# Changelog
+
+- v1.0.1
+  - Support RTL languages
+  - Add boxes for references
+  - Add custom score sliders for multi-dimensional evaluation
+  - Make instructions customizable and protocol-dependent
+  - Support custom sliders
+  - Purge/reset whole tasks from dashboard
+  - Fix resetting individual users in single-stream/dynamic
+  - Fix notification stacking
+  - Add campaigns from dashboard
+- v0.3.3
+  - Rename `doc_id` to `item_id`
+  - Add Typst, LaTeX, and PDF export for model ranking tables. Hide them by default.
+  - Add dynamic assignment type with contrastive model comparison
+  - Add `instructions_goodbye` field with variable substitution
+  - Add visual anchors at 33% and 66% on sliders
+  - Add German→English ESA tutorial with attention checks
+  - Validate document model consistency before shuffle
+  - Fix UI block on any interaction
+- v0.3.2
+  - Revert seeding of user IDs
+  - Set ESA (Error Span Annotation) as default
+  - Update server IP address configuration
+  - Show approximate alignment by default
+  - Unify pointwise and listwise interfaces into `basic`
+  - Refactor protocol configuration (breaking change)
+- v0.2.11
+  - Add comment field in settings panel
+  - Add `score_gt` validation for listwise comparisons
+  - Add Content-Disposition headers for proper download filenames
+  - Add model results display to dashboard with rankings
+  - Add campaign file structure validation
+  - Purge command now unlinks assets
+- v0.2.6
+  - Add frozen annotation links feature for view-only mode
+  - Add word-level annotation mode toggle for error spans
+  - Add `[missing]` token support
+  - Improve frontend speed and cleanup toolboxes on item load
+  - Host assets via symlinks
+  - Add validation threshold for success/fail tokens
+  - Implement reset masking for annotations
+  - Allow pre-defined user IDs and tokens in campaign data
+- v0.1.1
+  - Set server defaults and add VM launch scripts
+  - Add warning dialog when navigating away with unsaved work
+  - Add tutorial validation support for pointwise and listwise
+  - Add ability to preview existing annotations via progress bar
+  - Add support for ESA<sup>AI</sup> pre-filled error_spans
+  - Rename pairwise to listwise and update layout
+  - Implement single-stream assignment type
+- v0.0.3
+  - Support multimodal inputs and outputs
+  - Add dashboard
+  - Implement ESA (Error Span Annotation) and MQM support
+

{pearmut-1.0.0 → pearmut-1.0.1}/README.md

@@ -1,14 +1,7 @@
-# Pearmut 🍐
+# 🍐Pearmut     [![PyPi version](https://badgen.net/pypi/v/pearmut/)](https://pypi.org/project/pearmut) [![PyPI download/month](https://img.shields.io/pypi/dm/pearmut.svg)](https://pypi.python.org/pypi/pearmut/) [![PyPi license](https://badgen.net/pypi/license/pearmut/)](https://pypi.org/project/pearmut/) [![build status](https://github.com/zouharvi/pearmut/actions/workflows/test.yml/badge.svg)](https://github.com/zouharvi/pearmut/actions/workflows/test.yml)
 
 **Platform for Evaluation and Reviewing of Multilingual Tasks**: Evaluate model outputs for translation and NLP tasks with support for multimodal data (text, video, audio, images) and multiple annotation protocols ([DA](https://aclanthology.org/N15-1124/), [ESA](https://aclanthology.org/2024.wmt-1.131/), [ESA<sup>AI</sup>](https://aclanthology.org/2025.naacl-long.255/), [MQM](https://doi.org/10.1162/tacl_a_00437), and more!).
 
-[![PyPi version](https://badgen.net/pypi/v/pearmut/)](https://pypi.org/project/pearmut)
-&nbsp;
-[![PyPI download/month](https://img.shields.io/pypi/dm/pearmut.svg)](https://pypi.python.org/pypi/pearmut/)
-&nbsp;
-[![PyPi license](https://badgen.net/pypi/license/pearmut/)](https://pypi.org/project/pearmut/)
-&nbsp;
-[![build status](https://github.com/zouharvi/pearmut/actions/workflows/test.yml/badge.svg)](https://github.com/zouharvi/pearmut/actions/workflows/test.yml)
 
 <img width="1000" alt="Screenshot of ESA/MQM interface" src="https://github.com/user-attachments/assets/71334238-300b-4ffc-b777-7f3c242b1630" />
 
@@ -31,6 +24,8 @@
 - [Terminology](#terminology)
 - [Development](#development)
 - [Citation](#citation)
+- [Changelog](#changelog)
+
 
 ## Quick Start
 
@@ -90,7 +85,9 @@ Campaigns are defined in JSON files (see [examples/](examples/)). The simplest c
 }
 ```
 
-Each item has to have `src` (string) and `tgt` (dictionary from model names to strings, even for a single model evaluation).
+Each item has to have `tgt` (dictionary from model names to strings, even for a single model evaluation).
+Optionally, you can also include `src` (source string) and/or `ref` (reference string).
+If neither `src` nor `ref` is provided, only the model outputs will be displayed.
 For full Pearmut functionality (e.g. automatic statistical analysis), add `item_id` as well.
 Any other keys that you add will simply be stored in the logs.
 
@@ -124,6 +121,40 @@ The `shuffle` parameter in campaign `info` controls this behavior:
 }
 ```
 
+### Custom Score Sliders
+
+For multi-dimensional evaluation tasks (e.g., assessing fluency on a Likert scale), you can define custom sliders with specific ranges and steps:
+
+```python
+{
+  "info": {
+    "assignment": "task-based",
+    "protocol": "ESA",
+    "sliders": [
+      {"name": "Fluency", "min": 0, "max": 5, "step": 1},
+      {"name": "Adequacy", "min": 0, "max": 100, "step": 1}
+    ]
+  },
+  "campaign_id": "my_campaign",
+  "data": [...]
+}
+```
+
+When `sliders` is specified, only the custom sliders are shown. Each slider must have `name`, `min`, `max`, and `step` properties. All sliders must be answered before proceeding.
+
+### Custom Instructions
+
+Set campaign-level instructions using the `instructions` field in `info` (supports HTML).
+Instructions default to protocol-specific ones (DA: scoring, ESA: error spans + scoring, MQM: error spans + categories + scoring).
+```python
+{
+  "info": {
+    "protocol": "DA",
+    "instructions": "Rate translation quality on a 0-100 scale.<br>Pay special attention to document-level phenomena."
+  }
+}
+```
+
 ### Pre-filled Error Spans (ESA<sup>AI</sup>)
 
 Include `error_spans` to pre-fill annotations that users can review, modify, or delete:
@@ -242,7 +273,7 @@ All items must contain outputs from all models for this assignment type to work
 **How it works:**
 1. Initial phase: Each model gets `dynamic_first` annotations with fully random contrastive evaluation
 2. Dynamic phase: After the initial phase, top `dynamic_top` models (by average score) are identified
-3. Contrastive evaluatoin: From the top N models, `dynamic_contrastive_models` models are randomly selected for each item
+3. Contrastive evaluation: From the top N models, `dynamic_contrastive_models` models are randomly selected for each item
 4. Item prioritization: Items with the least annotations for the selected models are prioritized
 5. Backoff: With probability `dynamic_backoff`, uniform random selection is used instead to maintain exploration
 
@@ -268,6 +299,7 @@ The `users` field accepts:
 }
 ```
 
+
 ### Multimodal Annotations
 
 Support for HTML-compatible elements (YouTube embeds, `<video>` tags, images). Ensure elements are pre-styled. See [examples/multimodal.json](examples/multimodal.json).
@@ -397,3 +429,61 @@ If you use this work in your paper, please cite as following.
 ```
 
 Contributions are welcome! Please reach out to [Vilém Zouhar](mailto:vilem.zouhar@gmail.com).
+
+# Changelog
+
+- v1.0.1
+  - Support RTL languages
+  - Add boxes for references
+  - Add custom score sliders for multi-dimensional evaluation
+  - Make instructions customizable and protocol-dependent
+  - Support custom sliders
+  - Purge/reset whole tasks from dashboard
+  - Fix resetting individual users in single-stream/dynamic
+  - Fix notification stacking
+  - Add campaigns from dashboard
+- v0.3.3
+  - Rename `doc_id` to `item_id`
+  - Add Typst, LaTeX, and PDF export for model ranking tables. Hide them by default.
+  - Add dynamic assignment type with contrastive model comparison
+  - Add `instructions_goodbye` field with variable substitution
+  - Add visual anchors at 33% and 66% on sliders
+  - Add German→English ESA tutorial with attention checks
+  - Validate document model consistency before shuffle
+  - Fix UI block on any interaction
+- v0.3.2
+  - Revert seeding of user IDs
+  - Set ESA (Error Span Annotation) as default
+  - Update server IP address configuration
+  - Show approximate alignment by default
+  - Unify pointwise and listwise interfaces into `basic`
+  - Refactor protocol configuration (breaking change)
+- v0.2.11
+  - Add comment field in settings panel
+  - Add `score_gt` validation for listwise comparisons
+  - Add Content-Disposition headers for proper download filenames
+  - Add model results display to dashboard with rankings
+  - Add campaign file structure validation
+  - Purge command now unlinks assets
+- v0.2.6
+  - Add frozen annotation links feature for view-only mode
+  - Add word-level annotation mode toggle for error spans
+  - Add `[missing]` token support
+  - Improve frontend speed and cleanup toolboxes on item load
+  - Host assets via symlinks
+  - Add validation threshold for success/fail tokens
+  - Implement reset masking for annotations
+  - Allow pre-defined user IDs and tokens in campaign data
+- v0.1.1
+  - Set server defaults and add VM launch scripts
+  - Add warning dialog when navigating away with unsaved work
+  - Add tutorial validation support for pointwise and listwise
+  - Add ability to preview existing annotations via progress bar
+  - Add support for ESA<sup>AI</sup> pre-filled error_spans
+  - Rename pairwise to listwise and update layout
+  - Implement single-stream assignment type
+- v0.0.3
+  - Support multimodal inputs and outputs
+  - Add dashboard
+  - Implement ESA (Error Span Annotation) and MQM support
+

{pearmut-1.0.0 → pearmut-1.0.1}/pearmut.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pearmut
-Version: 1.0.0
+Version: 1.0.1
 Summary: A tool for evaluation of model outputs, primarily MT.
 Author-email: Vilém Zouhar <vilem.zouhar@gmail.com>
 License: MIT
@@ -19,17 +19,10 @@ Provides-Extra: dev
 Requires-Dist: pytest; extra == "dev"
 Dynamic: license-file
 
-# Pearmut 🍐
+# 🍐Pearmut &nbsp; &nbsp; [![PyPi version](https://badgen.net/pypi/v/pearmut/)](https://pypi.org/project/pearmut) [![PyPI download/month](https://img.shields.io/pypi/dm/pearmut.svg)](https://pypi.python.org/pypi/pearmut/) [![PyPi license](https://badgen.net/pypi/license/pearmut/)](https://pypi.org/project/pearmut/) [![build status](https://github.com/zouharvi/pearmut/actions/workflows/test.yml/badge.svg)](https://github.com/zouharvi/pearmut/actions/workflows/test.yml)
 
 **Platform for Evaluation and Reviewing of Multilingual Tasks**: Evaluate model outputs for translation and NLP tasks with support for multimodal data (text, video, audio, images) and multiple annotation protocols ([DA](https://aclanthology.org/N15-1124/), [ESA](https://aclanthology.org/2024.wmt-1.131/), [ESA<sup>AI</sup>](https://aclanthology.org/2025.naacl-long.255/), [MQM](https://doi.org/10.1162/tacl_a_00437), and more!).
 
-[![PyPi version](https://badgen.net/pypi/v/pearmut/)](https://pypi.org/project/pearmut)
-&nbsp;
-[![PyPI download/month](https://img.shields.io/pypi/dm/pearmut.svg)](https://pypi.python.org/pypi/pearmut/)
-&nbsp;
-[![PyPi license](https://badgen.net/pypi/license/pearmut/)](https://pypi.org/project/pearmut/)
-&nbsp;
-[![build status](https://github.com/zouharvi/pearmut/actions/workflows/test.yml/badge.svg)](https://github.com/zouharvi/pearmut/actions/workflows/test.yml)
 
 <img width="1000" alt="Screenshot of ESA/MQM interface" src="https://github.com/user-attachments/assets/71334238-300b-4ffc-b777-7f3c242b1630" />
 
@@ -52,6 +45,8 @@ Dynamic: license-file
 - [Terminology](#terminology)
 - [Development](#development)
 - [Citation](#citation)
+- [Changelog](#changelog)
+
 
 ## Quick Start
 
@@ -111,7 +106,9 @@ Campaigns are defined in JSON files (see [examples/](examples/)). The simplest c
 }
 ```
 
-Each item has to have `src` (string) and `tgt` (dictionary from model names to strings, even for a single model evaluation).
+Each item has to have `tgt` (dictionary from model names to strings, even for a single model evaluation).
+Optionally, you can also include `src` (source string) and/or `ref` (reference string).
+If neither `src` nor `ref` is provided, only the model outputs will be displayed.
 For full Pearmut functionality (e.g. automatic statistical analysis), add `item_id` as well.
 Any other keys that you add will simply be stored in the logs.
 
@@ -145,6 +142,40 @@ The `shuffle` parameter in campaign `info` controls this behavior:
 }
 ```
 
+### Custom Score Sliders
+
+For multi-dimensional evaluation tasks (e.g., assessing fluency on a Likert scale), you can define custom sliders with specific ranges and steps:
+
+```python
+{
+  "info": {
+    "assignment": "task-based",
+    "protocol": "ESA",
+    "sliders": [
+      {"name": "Fluency", "min": 0, "max": 5, "step": 1},
+      {"name": "Adequacy", "min": 0, "max": 100, "step": 1}
+    ]
+  },
+  "campaign_id": "my_campaign",
+  "data": [...]
+}
+```
+
+When `sliders` is specified, only the custom sliders are shown. Each slider must have `name`, `min`, `max`, and `step` properties. All sliders must be answered before proceeding.
+
+### Custom Instructions
+
+Set campaign-level instructions using the `instructions` field in `info` (supports HTML).
+Instructions default to protocol-specific ones (DA: scoring, ESA: error spans + scoring, MQM: error spans + categories + scoring).
+```python
+{
+  "info": {
+    "protocol": "DA",
+    "instructions": "Rate translation quality on a 0-100 scale.<br>Pay special attention to document-level phenomena."
+  }
+}
+```
+
 ### Pre-filled Error Spans (ESA<sup>AI</sup>)
 
 Include `error_spans` to pre-fill annotations that users can review, modify, or delete:
@@ -263,7 +294,7 @@ All items must contain outputs from all models for this assignment type to work
 **How it works:**
 1. Initial phase: Each model gets `dynamic_first` annotations with fully random contrastive evaluation
 2. Dynamic phase: After the initial phase, top `dynamic_top` models (by average score) are identified
-3. Contrastive evaluatoin: From the top N models, `dynamic_contrastive_models` models are randomly selected for each item
+3. Contrastive evaluation: From the top N models, `dynamic_contrastive_models` models are randomly selected for each item
 4. Item prioritization: Items with the least annotations for the selected models are prioritized
 5. Backoff: With probability `dynamic_backoff`, uniform random selection is used instead to maintain exploration
 
@@ -289,6 +320,7 @@ The `users` field accepts:
 }
 ```
 
+
 ### Multimodal Annotations
 
 Support for HTML-compatible elements (YouTube embeds, `<video>` tags, images). Ensure elements are pre-styled. See [examples/multimodal.json](examples/multimodal.json).
@@ -418,3 +450,61 @@ If you use this work in your paper, please cite as following.
 ```
 
 Contributions are welcome! Please reach out to [Vilém Zouhar](mailto:vilem.zouhar@gmail.com).
+
+# Changelog
+
+- v1.0.1
+  - Support RTL languages
+  - Add boxes for references
+  - Add custom score sliders for multi-dimensional evaluation
+  - Make instructions customizable and protocol-dependent
+  - Support custom sliders
+  - Purge/reset whole tasks from dashboard
+  - Fix resetting individual users in single-stream/dynamic
+  - Fix notification stacking
+  - Add campaigns from dashboard
+- v0.3.3
+  - Rename `doc_id` to `item_id`
+  - Add Typst, LaTeX, and PDF export for model ranking tables. Hide them by default.
+  - Add dynamic assignment type with contrastive model comparison
+  - Add `instructions_goodbye` field with variable substitution
+  - Add visual anchors at 33% and 66% on sliders
+  - Add German→English ESA tutorial with attention checks
+  - Validate document model consistency before shuffle
+  - Fix UI block on any interaction
+- v0.3.2
+  - Revert seeding of user IDs
+  - Set ESA (Error Span Annotation) as default
+  - Update server IP address configuration
+  - Show approximate alignment by default
+  - Unify pointwise and listwise interfaces into `basic`
+  - Refactor protocol configuration (breaking change)
+- v0.2.11
+  - Add comment field in settings panel
+  - Add `score_gt` validation for listwise comparisons
+  - Add Content-Disposition headers for proper download filenames
+  - Add model results display to dashboard with rankings
+  - Add campaign file structure validation
+  - Purge command now unlinks assets
+- v0.2.6
+  - Add frozen annotation links feature for view-only mode
+  - Add word-level annotation mode toggle for error spans
+  - Add `[missing]` token support
+  - Improve frontend speed and cleanup toolboxes on item load
+  - Host assets via symlinks
+  - Add validation threshold for success/fail tokens
+  - Implement reset masking for annotations
+  - Allow pre-defined user IDs and tokens in campaign data
+- v0.1.1
+  - Set server defaults and add VM launch scripts
+  - Add warning dialog when navigating away with unsaved work
+  - Add tutorial validation support for pointwise and listwise
+  - Add ability to preview existing annotations via progress bar
+  - Add support for ESA<sup>AI</sup> pre-filled error_spans
+  - Rename pairwise to listwise and update layout
+  - Implement single-stream assignment type
+- v0.0.3
+  - Support multimodal inputs and outputs
+  - Add dashboard
+  - Implement ESA (Error Span Annotation) and MQM support
+

{pearmut-1.0.0 → pearmut-1.0.1}/pearmut.egg-info/SOURCES.txt

@@ -10,6 +10,7 @@ pearmut.egg-info/top_level.txt
 server/app.py
 server/assignment.py
 server/cli.py
+server/constants.py
 server/results_export.py
 server/utils.py
 server/static/basic.bundle.js

{pearmut-1.0.0 → pearmut-1.0.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "pearmut"
-version = "1.0.0"
+version = "1.0.1"
 description = "A tool for evaluation of model outputs, primarily MT."
 readme = "README.md"
 license = { text = "MIT" }

{pearmut-1.0.0 → pearmut-1.0.1}/server/app.py

@@ -280,6 +280,79 @@ async def _reset_task(request: ResetTaskRequest):
     return response
 
 
+class PurgeCampaignRequest(BaseModel):
+    campaign_id: str
+    token: str
+
+
+@app.post("/purge-campaign")
+async def _purge_campaign(request: PurgeCampaignRequest):
+    global progress_data, tasks_data
+    
+    campaign_id = request.campaign_id
+    token = request.token
+
+    if campaign_id not in progress_data:
+        return JSONResponse(content="Unknown campaign ID", status_code=400)
+    if token != tasks_data[campaign_id]["token"]:
+        return JSONResponse(content="Invalid token", status_code=400)
+
+    # Unlink assets if they exist
+    destination = tasks_data[campaign_id].get("info", {}).get("assets", {}).get("destination")
+    if destination:
+        symlink_path = f"{ROOT}/data/{destination}".rstrip("/")
+        if os.path.islink(symlink_path):
+            os.remove(symlink_path)
+    
+    # Remove task file
+    task_file = f"{ROOT}/data/tasks/{campaign_id}.json"
+    if os.path.exists(task_file):
+        os.remove(task_file)
+    
+    # Remove output file
+    output_file = f"{ROOT}/data/outputs/{campaign_id}.jsonl"
+    if os.path.exists(output_file):
+        os.remove(output_file)
+    
+    # Remove from in-memory data structures
+    del tasks_data[campaign_id]
+    del progress_data[campaign_id]
+    
+    # Save updated progress data
+    save_progress_data(progress_data)
+    
+    return JSONResponse(content="ok", status_code=200)
+
+
+class AddCampaignRequest(BaseModel):
+    campaign_data: dict[str, Any]
+
+
+@app.post("/add-campaign")
+async def _add_campaign(request: AddCampaignRequest):
+    global progress_data, tasks_data
+    
+    from .cli import _add_single_campaign
+    
+    try:
+        server = f"{os.environ.get('PEARMUT_SERVER_URL', 'http://localhost:8001')}"
+        _add_single_campaign(request.campaign_data, overwrite=False, server=server)
+        
+        campaign_id = request.campaign_data['campaign_id']
+        with open(f"{ROOT}/data/tasks/{campaign_id}.json", "r") as f:
+            tasks_data[campaign_id] = json.load(f)
+        
+        progress_data = load_progress_data(warn=None)
+        
+        return JSONResponse(content={
+            "status": "ok",
+            "campaign_id": campaign_id,
+            "token": tasks_data[campaign_id]["token"]
+        }, status_code=200)
+    except Exception as e:
+        return JSONResponse(content={"error": str(e)}, status_code=400)
+
+
 @app.get("/download-annotations")
 async def _download_annotations(
     campaign_id: list[str] = Query(),