Update to Pendar

Changed files

Choose a file to see its text changes. Binary or large files show metadata only.

.app.json →

.codex-plugin/plugin.json →

.mcp.json →

assets/pendar-logo.png →

skills/audit-a-bibliography/SKILL.md →

skills/audit-a-bibliography/agents/openai.yaml →

skills/audit-a-bibliography/assets/pendar-logo.png →

skills/choose-a-venue/SKILL.md →

skills/choose-a-venue/agents/openai.yaml →

skills/choose-a-venue/assets/pendar-logo.png →

skills/map-a-field/SKILL.md →

skills/map-a-field/agents/openai.yaml →

skills/map-a-field/assets/pendar-logo.png →

skills/map-my-research/SKILL.md →

skills/map-my-research/agents/openai.yaml →

skills/map-my-research/assets/pendar-logo.png →

skills/profile-a-researcher/SKILL.md →

skills/profile-a-researcher/agents/openai.yaml →

skills/profile-a-researcher/assets/pendar-logo.png →

skills/survey-a-research-field/SKILL.md →

skills/survey-a-research-field/assets/pendar-logo.png →

skills/trace-citation-impact/SKILL.md →

skills/trace-citation-impact/agents/openai.yaml →

skills/trace-citation-impact/assets/pendar-logo.png →

skills/verify-citations/SKILL.md →

skills/verify-citations/assets/pendar-logo.png →

.codex-plugin/plugin.json

− removed · + added

--- before
+++ after
@@ -1,27 +1,89 @@
 {
-  "apps": "./.app.json",
+  "name": "app-6a8399a0bccc819185d57048383c915c",
+  "version": "4.0.0",
+  "description": "Go beyond finding papers. Discover how research fields are structured, how ideas connect, and where science is heading.\n\nPendar helps you explore and analyze scientific research using evidence from the published literature\u2014not guesses or generic AI summaries.\n\nUnderstand your research. Tell Pendar the name you publish under and uncover promising areas you haven't explored yet. Discover your main research themes, see how they are evolving and identify the fields building on your work.\n\nExplore a research field. See how a field is organized into different themes, how those themes connect to neighboring disciplines, and which areas are growing or gaining momentum.\n\nMeasure the size and direction of a field. Find out how many papers have been published, whether research activity is growing or declining, which subfields are advancing fastest, who funds the work, and which papers are worth reading first.\n\nTrace the influence of a paper. See which research areas have built on a publication, how its ideas have spread, and whether its citation activity is increasing over time.\n\nDiscover journals and researchers. Find journals publishing in a particular research area and explore the work of colleagues, potential collaborators, or prospective advisors.\n\nVerify your references. Paste a bibliography to check each reference against the published record. Identify incorrect citations and retracted papers. Pendar verifies the references you provide rather than silently substituting similar-looking publications.\n\nDesigned for graduate students, postdocs, professors, and researchers planning their next paper, grant, collaboration, or career move.\n\nGrounded in published research. Figures are counted across the indexed literature rather than estimated from a small sample.\n\nExplore more at pendar.ai, including deep research and comprehensive research analysis.",
   "author": {
     "name": "Pendarlabs"
   },
-  "description": "Pendar gives you the shape of a research field, not a summary of it.\n\nPoint it at a topic and it reports how many papers exist, whether output is growing or flattening, which disciplines and countries produce them, and which papers to read first. Every figure is counted across the whole corpus rather than estimated from a sample, so you can plan around it.\n\nPoint it at a paper and it traces what happened next: whether citations are rising or fading, which fields picked the work up, whether industry or academia is building on it, and which authors cite it most. That is how you tell a paper quietly becoming foundational from one that peaked years ago.\n\nPaste a bibliography and it checks each reference against the published record: whether the paper exists, whether the citation describes it accurately, and whether it has been retracted. A reference it cannot find is reported as unverified rather than swapped for the nearest real paper, which matters when the citations came from a model.\n\nEvery tool is read-only. Pendar looks things up and reports what it finds.",
+  "skills": "./skills",
   "interface": {
-    "capabilities": [],
-    "category": "Scientific Research",
-    "defaultPrompt": [
-      "Survey CRISPR gene editing: how many papers exist, is output growing, and what should I read first?",
-      "For DOI 10.1038/s41586-021-03819-2, which fields adopted it and is industry or academia citing it more?",
-      "Check whether the papers in my reference list are real, correctly cited, and not retracted"
-    ],
-    "developerName": "Pendarlabs",
     "displayName": "Pendar",
-    "longDescription": "Pendar gives you the shape of a research field, not a summary of it.\n\nPoint it at a topic and it reports how many papers exist, whether output is growing or flattening, which disciplines and countries produce them, and which papers to read first. Every figure is counted across the whole corpus rather than estimated from a sample, so you can plan around it.\n\nPoint it at a paper and it traces what happened next: whether citations are rising or fading, which fields picked the work up, whether industry or academia is building on it, and which authors cite it most. That is how you tell a paper quietly becoming foundational from one that peaked years ago.\n\nPaste a bibliography and it checks each reference against the published record: whether the paper exists, whether the citation describes it accurately, and whether it has been retracted. A reference it cannot find is reported as unverified rather than swapped for the nearest real paper, which matters when the citations came from a model.\n\nEvery tool is read-only. Pendar looks things up and reports what it finds.",
-    "privacyPolicyURL": "https://pendar.ai/privacy/",
-    "shortDescription": "Map & measure research fields",
+    "shortDescription": "The bigger picture of research",
+    "longDescription": "Go beyond finding papers. Discover how research fields are structured, how ideas connect, and where science is heading.\n\nPendar helps you explore and analyze scientific research using evidence from the published literature\u2014not guesses or generic AI summaries.\n\nUnderstand your research. Tell Pendar the name you publish under and uncover promising areas you haven't explored yet. Discover your main research themes, see how they are evolving and identify the fields building on your work.\n\nExplore a research field. See how a field is organized into different themes, how those themes connect to neighboring disciplines, and which areas are growing or gaining momentum.\n\nMeasure the size and direction of a field. Find out how many papers have been published, whether research activity is growing or declining, which subfields are advancing fastest, who funds the work, and which papers are worth reading first.\n\nTrace the influence of a paper. See which research areas have built on a publication, how its ideas have spread, and whether its citation activity is increasing over time.\n\nDiscover journals and researchers. Find journals publishing in a particular research area and explore the work of colleagues, potential collaborators, or prospective advisors.\n\nVerify your references. Paste a bibliography to check each reference against the published record. Identify incorrect citations and retracted papers. Pendar verifies the references you provide rather than silently substituting similar-looking publications.\n\nDesigned for graduate students, postdocs, professors, and researchers planning their next paper, grant, collaboration, or career move.\n\nGrounded in published research. Figures are counted across the indexed literature rather than estimated from a small sample.\n\nExplore more at pendar.ai, including deep research and comprehensive research analysis.",
+    "developerName": "Pendarlabs",
+    "category": "Scientific Research",
+    "capabilities": [],
+    "websiteURL": "https://pendar.ai",
     "supportURL": "https://pendar.ai/support/",
+    "privacyPolicyURL": "https://pendar.ai/privacy/",
     "termsOfServiceURL": "https://pendar.ai/terms/",
-    "websiteURL": "https://pendar.ai"
+    "defaultPrompt": [
+      "Where does my research stand in its field, and which growing areas nearby am I missing?",
+      "Since COVID, where is wastewater-based epidemiology heading, and which nearby fields are growing fastest?",
+      "Survey microplastics research: how big is it, is it growing, and who funds it?"
+    ],
+    "composerIcon": "./assets/pendar-logo.png",
+    "logo": "./assets/pendar-logo.png"
   },
-  "name": "app-6a8399a0bccc819185d57048383c915c",
-  "skills": "./skills",
-  "version": "2.0.0"
-}
\ No newline at end of file
+  "extensions": {
+    "com.openai": {
+      "review": {
+        "demo_recording_url": "https://youtu.be/72aRxih-t_E",
+        "test_cases": {
+          "positive": [
+            {
+              "description": "Map where a researcher's work stands: home themes, who builds on it, and close, growing themes they have not published in.",
+              "prompt": "I publish as Dino Di Carlo at UCLA. Where does my research stand, and which growing areas near it am I missing?",
+              "tools_triggered": "my_position",
+              "expected_behavior": "One my_position call. Home themes with momentum, who cites the work, and blind spots with their reasons. No ranking of the researcher."
+            },
+            {
+              "description": "Map a field's neighbours: which adjacent fields are growing, and which the field is moving toward.",
+              "prompt": "Which fields next to solid-state battery electrolytes are growing fastest, and which is the field moving toward?",
+              "tools_triggered": "field_map",
+              "expected_behavior": "One field_map call. Core themes, the converging neighbour, and other growing and fading neighbours. Counts come from the tool."
+            },
+            {
+              "description": "Size up a research field: how many papers exist, which sub-areas are growing, and who funds the work.",
+              "prompt": "Survey CRISPR gene editing: how many papers exist, which sub-areas are growing fastest, and who funds the work?",
+              "tools_triggered": "survey_literature",
+              "expected_behavior": "One survey_literature call. Paper count, growth trend, fastest-growing sub-areas, and funders. Counts come from the tool."
+            },
+            {
+              "description": "Measure a paper's impact: which fields cite it now, and whether industry or academia is building on it.",
+              "prompt": "For DOI 10.1038/s41586-021-03819-2, which fields cite it more lately, and is industry or academia building on it?",
+              "tools_triggered": "graph_hopper",
+              "expected_behavior": "One graph_hopper call. Citing-paper counts, which fields grew, and the academia versus industry split."
+            },
+            {
+              "description": "Summarize a named researcher's published work, using the institution to identify the right person.",
+              "prompt": "What does Heinrich Jaeger at the University of Chicago work on, and has his focus shifted?",
+              "tools_triggered": "profile_researcher",
+              "expected_behavior": "One profile_researcher call. Main topics, whether the focus shifted, and no contact details."
+            }
+          ],
+          "negative": [
+            {
+              "description": "Explain how a method works. No lookup is needed.",
+              "prompt": "Explain how AlphaFold predicts protein structure from an amino acid sequence."
+            },
+            {
+              "description": "Reformat a reference into a citation style. The task is formatting, not checking the citation.",
+              "prompt": "Reformat this reference into APA style: Jumper et al., AlphaFold, Nature 596, 583 (2021)."
+            },
+            {
+              "description": "Draft a cover letter. The task is writing, not a literature lookup.",
+              "prompt": "Draft a cover letter to the editor of Nature Communications for my battery paper."
+            }
+          ]
+        },
+        "commerce": false,
+        "commerce_description": "Pendar sells nothing inside ChatGPT. Each tool call uses credits from the user's Pendar account; new accounts start with free credits, and further credits are bought on pendar.ai."
+      },
+      "publication": {
+        "release_notes": "4.0.0 maps research. Two new tools: my_position places a researcher's published work among the themes around it, with home themes and their momentum, who builds on the work, and blind spots (close, growing themes they have not published in); field_map maps a field's core themes and its neighbours, with which are growing and which the field is moving toward. Two new skills, map-my-research and map-a-field, read and draw these maps. Citation checking stays: verify_paper and audit_bibliography read more citation styles, accept arXiv and PubMed IDs, report a cited title that differs from the published one, and say when a lookup could not be completed instead of calling a paper missing."
+      }
+    }
+  },
+  "apps": "./.app.json"
+}