Radar calibration
Scoring and Role Derivation
Denominator: In-cluster, N=22. Source: Manual-audit/CALIBRATION-TABLE-RADAR-VS-CODE-AI-CODING-AGENT-EXTENSIONS.md. Reader meaning: this page shows how accepted SC-024 scores produce public role labels without changing the underlying analysis.
Decision Rules
Scores use a 0-5 scale. Public labels use neutral Sector Radar vocabulary. Decision rules are applied in order.
| Priority | Condition | Public role | Reader meaning |
|---|---|---|---|
| 1 | HE <= 1 | Inactive Scaffold | Structurally present but little sustained engineering signal. |
| 2 | HE >= 4 AND AD >= 4 AND CM >= 3 | Category Leader | High engineering depth, public reach, and centroid alignment. |
| 3 | CM <= 2 AND HE >= 3 | Adjacent Infrastructure | Boundary-visible but not the default extension pattern. |
| 4 | DF >= 3 AND HE >= 2 | Differentiated Niche | Distinct sector position with enough engineering depth. |
| 5 | CM >= 4 AND DF <= 2 AND HE >= 2 | Strong Default | Recognizable default implementation with limited differentiation. |
| 6 | CM >= 3 AND DF <= 1 | Baseline Implementation | Close to the default with thin independent signal. |
Score Dimensions
| Dimension | Code | 0-5 scale anchor |
|---|---|---|
| Centroid Match | CM | 0=no centroid traits; 5=9-10 centroid traits present. |
| Differentiation | DF | 0=no observable differentiators; 5=5+ functional differentiators beyond centroid. |
| Human-Edge Strength | HE | 0=no sustained engineering; 5=deep, sustained, reviewable engineering. |
| Adoption Depth | AD | 0=no real adoption signal; 5=top-tier stars/forks/issues/use. |
| Claim/Implementation Integrity | CI | 0=claims unsupported; 5=major README/manifest claims trace to source evidence. |
In-Cluster Corpus Scores
Denominator: In-cluster, N=22. Source: calibration table score matrix. Reader meaning: each rule trace shows why a public role was assigned.
| # | Repo | Stars | CM | DF | HE | AD | CI | Public role | Rule trace |
|---|---|---|---|---|---|---|---|---|---|
| 1 | cline/cline | ~61,937 | 5 | 5 | 5 | 5 | 5 | Category Leader | HE=5, AD=5, CM=5. |
| 2 | continuedev/continue | ~33,234 | 5 | 5 | 5 | 5 | 5 | Category Leader | HE=5, AD=5, CM=5. |
| 3 | Kilo-Org/kilocode | ~19,362 | 5 | 4 | 5 | 5 | 4 | Category Leader | HE=5, AD=5, CM=5. |
| 4 | zgsm-ai/costrict | ~4,017 | 4 | 4 | 4 | 4 | 4 | Category Leader | HE=4, AD=4, CM=4. |
| 5 | codota/tabnine-vscode | ~1,440 | 3 | 3 | 4 | 4 | 4 | Category Leader | HE=4, AD=4, CM=3. |
| 6 | DeepMyst/Mysti | ~1,048 | 4 | 4 | 3 | 3 | 4 | Differentiated Niche | DF=4 and HE=3. |
| 7 | 10Nates/ollama-autocoder | ~151 | 3 | 1 | 2 | 2 | 3 | Baseline Implementation | CM=3 and DF=1. |
| 8 | Lianues/Lim-Code | ~140 | 4 | 2 | 3 | 2 | 3 | Strong Default | CM=4, DF=2, HE=3. |
| 9 | andrepimenta/claude-code-chat | ~1,049 | 4 | 3 | 3 | 4 | 4 | Differentiated Niche | DF=3 and HE=3. |
| 10 | adifyr/shadow-code | ~78 | 3 | 4 | 3 | 2 | 3 | Differentiated Niche | DF=4 and HE=3. |
| 11 | rusiaaman/chat.md | ~105 | 3 | 4 | 3 | 2 | 3 | Differentiated Niche | DF=4 and HE=3. |
| 12 | vscode-reborn-ai/vscode-reborn-ai | ~266 | 4 | 2 | 3 | 2 | 3 | Strong Default | CM=4, DF=2, HE=3. |
| 13 | cogflows/promptcode-vscode | ~15 | 2 | 3 | 3 | 1 | 3 | Adjacent Infrastructure | CM=2 and HE=3. |
| 14 | ZhouChaunge/DeepCopilot | ~15 | 3 | 1 | 2 | 1 | 3 | Baseline Implementation | CM=3 and DF=1. |
| 15 | Kile-Thomson/Rokket-GSD | ~15 | 4 | 2 | 2 | 1 | 3 | Strong Default | CM=4, DF=2, HE=2. |
| 16 | jasonjmcghee/claude-debugs-for-you | ~510 | 2 | 3 | 3 | 2 | 3 | Adjacent Infrastructure | CM=2 and HE=3. |
| 17 | Hive-Academy/ptah-extension | ~26 | 4 | 4 | 3 | 1 | 4 | Differentiated Niche | DF=4 and HE=3. |
| 18 | AizenvoltPrime/damocles | ~60 | 4 | 3 | 3 | 1 | 4 | Differentiated Niche | DF=3 and HE=3. |
| 19 | justimyhxu/claude-code-local-for-vscode | ~39 | 3 | 1 | 2 | 1 | 3 | Baseline Implementation | CM=3 and DF=1. |
| 20 | formulahendry/vscode-acp | ~244 | 2 | 4 | 3 | 2 | 4 | Adjacent Infrastructure | CM=2 and HE=3. |
| 21 | joaompfp/hermes-vscode | ~22 | 3 | 3 | 2 | 1 | 4 | Differentiated Niche | DF=3 and HE=2. |
| 22 | Harsh1210/openclaude-vscode | ~11 | 4 | 2 | 2 | 1 | 4 | Strong Default | CM=4, DF=2, HE=2. |
Boundary Judgments
Denominator: In-cluster, N=22. Source: calibration boundary and lineage register. Reader meaning: boundary notes explain why edge cases stay visible and how lineage/protocol evidence affects interpretation.
| Repo | Boundary decision | Lineage / protocol note | Public role |
|---|---|---|---|
| cline/cline | Accepted anchor with source-verified VS Code package and AI coding-agent headline. | Anchor project; later fork/lineage candidates are compared against the Cline-style pattern. | Category Leader |
| continuedev/continue | Accepted anchor with source-verified VS Code package and AI coding-agent headline. | Anchor project; not treated as Cline lineage. | Category Leader |
| Kilo-Org/kilocode | Accepted anchor and lineage representative with active package-backed source. | Lineage-bearing candidate retained as active source package with Cline-style relevance. | Category Leader |
| zgsm-ai/costrict | Accepted active VS Code AI coder extension. | No explicit lineage flag in SC-021; scored as independent accepted candidate. | Category Leader |
| DeepMyst/Mysti | Accepted active package-backed coding-agent extension. | Independent multi-agent niche. | Differentiated Niche |
| 10Nates/ollama-autocoder | Retained because source/package evidence clears the SC-021 gate. | Long-tail local-model implementation. | Baseline Implementation |
| Lianues/Lim-Code | Accepted active package-backed coding assistant. | Independent long-tail assistant. | Strong Default |
| adifyr/shadow-code | Accepted active package-backed extension, but not an anchor. | Workflow-niche implementation. | Differentiated Niche |
| rusiaaman/chat.md | Borderline retained; file-first chat sits near the edge but remains in-cluster. | VS Code package, MCP/tool use, and coding-agent relevance are source-backed. | Differentiated Niche |
| vscode-reborn-ai/vscode-reborn-ai | Accepted active package-backed coding assistant. | Independent assistant implementation. | Strong Default |
| cogflows/promptcode-vscode | Borderline retained; classified Adjacent Infrastructure inside the final corpus. | Prompt/apply workflow near the edge of the agent-extension boundary. | Adjacent Infrastructure |
| ZhouChaunge/DeepCopilot | Retained because source/package evidence clears the gate despite near-floor scale. | Low-star baseline implementation. | Baseline Implementation |
| jasonjmcghee/claude-debugs-for-you | Borderline retained; classified Adjacent Infrastructure because CM=2. | MCP-hybrid retained; not a full IDE coding-agent surface. | Adjacent Infrastructure |
| Hive-Academy/ptah-extension | Accepted active package-backed multi-agent extension. | Orchestration niche. | Differentiated Niche |
| AizenvoltPrime/damocles | Accepted active package-backed extension. | Claude-agent niche. | Differentiated Niche |
| justimyhxu/claude-code-local-for-vscode | Borderline lineage retained; public role is Baseline Implementation. | Lineage-bearing candidate retained as active source package patching Claude Code for local/remote mode. | Baseline Implementation |
| formulahendry/vscode-acp | Accepted as an agent-client extension surface. | ACP-client surface; explicitly distinguished from MCP-client cases. | Adjacent Infrastructure |
| joaompfp/hermes-vscode | Accepted low-star ACP agent-client extension. | ACP-client surface; not MCP-client and not Cline lineage. | Differentiated Niche |
| Harsh1210/openclaude-vscode | Retained despite near-star-floor advisory because source/package evidence clears the gate. | Long-tail near-floor accepted package. | Strong Default |
Cluster Composition
Denominator: In-cluster, N=22. Source: calibration composition table. Reader meaning: this table makes role distribution auditable from the same score source.
| Role | Count | Percentage | Repos |
|---|---|---|---|
| Category Leader | 5 | 22.7% | cline/cline, continuedev/continue, Kilo-Org/kilocode, zgsm-ai/costrict, codota/tabnine-vscode |
| Differentiated Niche | 7 | 31.8% | DeepMyst/Mysti, andrepimenta/claude-code-chat, adifyr/shadow-code, rusiaaman/chat.md, Hive-Academy/ptah-extension, AizenvoltPrime/damocles, joaompfp/hermes-vscode |
| Strong Default | 4 | 18.2% | Lianues/Lim-Code, vscode-reborn-ai/vscode-reborn-ai, Kile-Thomson/Rokket-GSD, Harsh1210/openclaude-vscode |
| Baseline Implementation | 3 | 13.6% | 10Nates/ollama-autocoder, ZhouChaunge/DeepCopilot, justimyhxu/claude-code-local-for-vscode |
| Adjacent Infrastructure | 3 | 13.6% | cogflows/promptcode-vscode, jasonjmcghee/claude-debugs-for-you, formulahendry/vscode-acp |
| Inactive Scaffold | 0 | 0.0% | None |
Sector Health Metrics
Denominator: In-cluster, N=22. Source: calibration health metrics and visual data table 4. Reader meaning: health metrics summarize the score matrix without replacing row-level evidence.
| Metric | Value | Reader meaning |
|---|---|---|
| Convergence Index | 3.55 / 5 | Moderate-high convergence around package-backed IDE agent surfaces. |
| Differentiation Index | 3.05 / 5 | Higher than Chrome Extension Starters because protocols, runtimes, and workflow shapes vary. |
| Human-Edge Index | 3.09 / 5 | Moderate; strong anchors plus a thin long tail. |
| Adoption Index | 2.36 / 5 | Top-heavy; Cline/Continue/Kilo dominate public reach. |
| Integrity Index | 3.64 / 5 | Generally source-backed, with several long-tail claim-depth limits. |
| Viability Rate | 72.7% (16/22) | Repos with HE >= 3. |
| Inactive Scaffold Rate | 0.0% (0/22) | No retained repo is classified as Inactive Scaffold. |
| Category Leakage Rate | 13.6% (3/22) | ACP/MCP/prompt tooling retained as boundary-visible adjacent infrastructure. |
Source Artifacts
Source map for this page. The calibration page surfaces accepted SC-024 scoring evidence and rule traces; it does not introduce new scores or role assignments.
| Published artifact | Source role | Visible metric families |
|---|---|---|
| Manual-audit/CALIBRATION-TABLE-RADAR-VS-CODE-AI-CODING-AGENT-EXTENSIONS.md | Canonical calibration table | Scores, rule traces, public roles, boundary judgments, health metrics. |
| Manual-audit/CENTROID-VS-CODE-AI-CODING-AGENT-EXTENSIONS.md | Centroid source | CM inputs and trait evidence. |
| Reports/VISUAL-DATA-VS-CODE-AI-CODING-AGENT-EXTENSIONS.md | Chart-ready aggregations | Composition, role averages, health metrics, and full score matrix. |
| Operations/Sector-Reports/SC-024/EVIDENCE-REFRESH.md | Evidence refresh context | Metadata refresh and source packet basis. |
| Operations/SECTOR-RADAR-DISPLAY-STANDARD.md | Display contract | Denominator, Source, Reader meaning, and correction-route requirements. |
Corrections
Found a stale repo, wrong source trace, or boundary classification issue? Send a correction to hello@diversum.dev with the repo name, displayed score or role, and source that should replace it.