Portfolio audit

Graf project / app / site audit

A boardroom-ready inventory of the workspaces we are carrying forward. Live filesystem stats are recomputed from the current machine, while purpose and output labels remain the curated descriptors that make each row readable at a glance.

Last refreshed 2026-08-31 10:36 UTC Cron jobs audited 7 active
Total roots
16
All inventory rows.
Live roots
16
Active project roots on this machine.
Git repos
10
Live repositories with git metadata.
Public URLs
6
Visible public/preview endpoints.
Cron jobs
7 active
Audited from ~/.hermes/cron/jobs.json.
Status legend
  • Workspace = non-git project folder or live working space.
  • Git = repo is present and tracked on its current branch.
Apps & product workspaces: 4 Sites & content: 4 Research & benchmarks: 6 Tooling / memory systems: 1 References / support: 1
Notes
Cron jobs and purpose
7 active enabled jobs loaded from ~/.hermes/cron/jobs.json.
JobPurposeScheduleState
BoilingFrogs subdomain health check
a0a866804f8b
Nightly BoilingFrogs subdomain health check and safe-repair job
0 1 * * * enabled
scheduled
Publishing dashboard data refresh
0ec2948337c2
Refresh the publishing monitoring dashboard data. Run `/home/gb001/apps/publishing-dashboard/scripts/update_dashboard.py`, verify `/home/gb001/apps/publishing-dashboard/public/data.json` exists and contains valid JSON, then finish with a one-line status. Do not ask questions. Do not modify KDP or publish any manuscript files
*/15 * * * * enabled
scheduled
Daily YouTube AI podcaster leaderboard refresh
cecc7965daed
Daily research/update job for the BoilingFrogs preview site
0 6 * * * enabled
scheduled
postgres-daily-health-audit
ea11fc9448fb
Run the safe read-only PostgreSQL health/audit using the deployed Hermes PostgreSQL ops helper. Do not perform any mutating database operation. Steps: (1) run `python3 /home/gb001/.hermes/postgres_ops/postgres_ops_cli.py registry`; (2) run `python3 /home/gb001/.hermes/postgres_ops/postgres_ops_cli.py health-check`; (3) run `python3 /home/gb001/.hermes/postgres_ops/postgres_ops_cli.py app-report --app ursa`; (4) write a concise Markdown report to `/home/gb001/.hermes/reports/postgres/daily-health-$(date -u +%Y%m%d).md` summarising registry apps, database connectivity, errors, connection counts, locks, and any follow-up needed. If psql authentication fails because `hermes_pg_ops` is not bootstrapped yet, state that clearly and do not retry with privileged/destructive commands
5 6 * * * enabled
scheduled
Adaptive email shadow cycle watchdog
9822c3b48295
Script-only job: adaptive-email-shadow-cycle.sh
0 9 * * * enabled
scheduled
Publish Boiling Frogs audit
7705cef5bbcf
Script-only job: publish-boilingfrogs-audit.sh
20 7 * * 1 enabled
scheduled
obsidian-drive-sync
5c2a49794e6c
Script-only job: sync_obsidian_index_to_drive.py
30 2 * * * enabled
scheduled
Project Purpose File structure / type Output type Website URL Recent activity Scale Status / notes
/home/gb001/apps/adaptive-inference-control-plane
Apps & product workspaces
Private Pioneer-style control plane for logging model traffic, collecting feedback, building training datasets, tracking model versions/evals, and preparing adaptive fine-tuning workflows.
README.md; package.json; package-lock.json; vite.config.js; index.html; .data/; artifacts/; backups/; +9 more
Top languages: JSON (1285f/2,860,107l) Markdown (500f/24,975l) Other (782f/15,512l) JavaScript (40f/5,384l)
Node/JavaScript web/API control plane with data pipeline, docs, migrations, and built assets.
0d · fresh · workspace
2620 files · 2,906,810 LOC · Large
workspace
  • Non-git workspace.
  • Scanned files: 2620
  • Filesystem scan includes generated/data folders: .data, data, dist, artifacts.
/home/gb001/apps/publishing-dashboard
Apps & product workspaces
Publishing dashboard with static public assets, Caddy config snapshots, and an update script.
README.md; public/; scripts/; caddy-config-before-basic-auth.json; caddy-config-before-publishing-dashboard.json
Top languages: JSON (3f/517l) Python (1f/121l) HTML (1f/51l) Markdown (1f/30l)
Static dashboard/content site plus Python refresh script.
publishing.boilingfrogs.ai 0d · fresh · workspace
6 files · 719 LOC · Tiny
workspace
  • Non-git workspace.
  • Scanned files: 6
/home/gb001/apps/realtime-voice-platform
Apps & product workspaces
Use-case-neutral realtime voice assistant platform for live conversation, retrieval, and post-turn analysis.
README.md; package.json; package-lock.json; backend/; deploy/; docs/; frontend/; runtime/; +2 more
Top languages: Other (3237f/36,754l) JSON (2159f/29,556l) JavaScript (52f/6,349l) React JSX (2f/907l)
Full-stack web app: Node backend + React frontend + SQL stores.
realtime.boilingfrogs.ai 44d · quiet · git
5464 files · 74,936 LOC · Medium
git on master
  • Tracked: 73
  • Last commit: 2026-07-18 14:24:12 +0000
  • No local changes.
/home/gb001/apps/todo-list
Apps & product workspaces
Lightweight todo-list app with PostgreSQL-backed hierarchy, category drill-down, and modal-based editing.
package.json; package-lock.json; vite.config.js; index.html; backend/; dist/; migrations/; src/; +1 more
Top languages: JSON (4f/2,861l) React JSX (2f/1,337l) JavaScript (3f/978l) CSS (2f/938l)
Vite/React frontend + Express/Postgres backend.
todo.boilingfrogs.ai 49d · quiet · workspace
15 files · 6,386 LOC · Small
workspace
  • Non-git workspace.
  • Scanned files: 15
/home/gb001/projects/boilingfrogs-audit-site
Sites & content
Inventory and publish the current BoilingFrogs portfolio audit, including the main table, cleanup summaries, and project detail pages.
dist/
Top languages: HTML (21f/2,833l)
Static HTML/CSS audit report site with mirrored publish root.
audit.boilingfrogs.ai 0d · fresh · workspace
21 files · 2,833 LOC · Small
workspace
  • Non-git workspace.
  • Scanned files: 21
/home/gb001/projects/boilingfrogs-site
Sites & content
BoilingFrogs.ai static site explaining AI changes and publishing model/news leaderboard content.
README.md; package.json; package-lock.json; astro.config.mjs; data/; db/; dist/; docs/; +8 more
Top languages: JSON (202f/1,047,872l) Other (137f/14,399l) Markdown (48f/3,323l) CSS (7f/2,737l)
Astro static site with data-refresh pipelines and generated content.
boilingfrogs.ai 18d · quiet · git
462 files · 1,072,927 LOC · Large
git on feat/nightly-publication-readiness-pass
  • Tracked: 321
  • Last commit: 2026-08-13 01:03:27 +0000
  • No local changes.
/home/gb001/projects/creativebench-site
Sites & content
CreativeBench is a single end-to-end manuscript-evaluation and publishing platform for agents, authors, editors, and paying readers.
package.json; package-lock.json; astro.config.mjs; dist/; docs/; migrations/; ops/; public/; +4 more
Top languages: JSON (2f/4,475l) Other (17f/1,039l) CSS (2f/986l) JavaScript (12f/857l)
Astro site with server/ops/migrations support files.
creativebench.boilingfrogs.ai 8d · active · git
50 files · 8,108 LOC · Small
git on main
  • Tracked: 38
  • Last commit: 2026-08-23 00:18:49 +0000
  • No local changes.
/home/gb001/projects/new-manuscript-series
Sites & content
New fiction series manuscript workspace with book, bible, and source materials.
README.md; book-01/; docs/; series-bible/; source/
Top languages: Markdown (3f/46l)
Markdown/content publishing assets.
44d · quiet · workspace
3 files · 46 LOC · Tiny
workspace
  • Non-git workspace.
  • Scanned files: 3
/home/gb001/research_bench/arena-hard-auto
Research & benchmarks
Automation harness for Arena-Hard model evaluation and judging.
README.md; requirements.txt; BenchBuilder/; config/; data/; leaderboard/; misc/; utils/; +6 more
Top languages: Other (17f/9,953l) Python (15f/3,859l) Markdown (3f/558l) YAML (6f/429l)
Benchmark/evaluation harness with browser QA helpers and data.
436d · stale · git
45 files · 14,885 LOC · Small
git on main
  • Tracked: 47
  • Last commit: 2025-06-21 04:44:12 +0000
  • No local changes.
/home/gb001/research_bench/creative-writing-bench
Research & benchmarks
Creative-writing benchmark with rubric scoring and Elo-style comparisons.
README.md; requirements.txt; core/; data/; results/; utils/; .env.example; creative_bench_runs.zip; +4 more
Top languages: JSON (6f/4,052,049l) HTML (31f/437,989l) Python (14f/4,306l) Other (3f/3,602l)
Benchmark/evaluation harness plus result HTML/JSON corpus.
92d · stale · git
71 files · 4,499,259 LOC · Large
git on main
  • Tracked: 351
  • Last commit: 2026-05-31 00:46:36 +0000
  • No local changes.
/home/gb001/research_bench/FastChat
Research & benchmarks
Open platform for training, serving, and evaluating LLM chatbots.
README.md; pyproject.toml; fastchat/; format.sh; LICENSE; theme.json
Top languages: Python (55f/20,383l) Other (3f/686l) Markdown (2f/442l) Shell (1f/71l)
Python package/research benchmark and serving stack.
455d · stale · git
65 files · 21,682 LOC · Medium
git on main
  • Tracked: 219
  • Last commit: 2025-06-02 15:22:03 +0000
  • No local changes.
/home/gb001/research_bench/longform-writing-bench
Research & benchmarks
Long-form writing benchmark for planning, revision, and story construction.
README.md; requirements.txt; core/; data/; results/; utils/; .env.example; generate_results_html.ipynb; +1 more
Top languages: JSON (7f/1,848,099l) HTML (69f/1,089,874l) Other (3f/4,682l) Python (8f/3,248l)
Benchmark/evaluation harness plus result HTML/JSON corpus.
309d · stale · git
114 files · 2,947,302 LOC · Large
git on main
  • Tracked: 392
  • Last commit: 2025-10-26 00:55:18 +0000
  • No local changes.
/home/gb001/research_bench/WildBench
Research & benchmarks
Benchmarking LLMs with challenging tasks from real users in the wild.
README.md; requirements.txt; docs/; eval_results/; evaluation/; leaderboard/; scripts/; src/; +3 more
Top languages: JSON (347f/7,957,147l) CSS (20f/38,367l) JavaScript (28f/33,104l) Python (23f/5,265l)
Benchmark/evaluation harness with leaderboard + eval result corpus.
760d · stale · git
504 files · 8,038,739 LOC · Large
git on main
  • Tracked: 535
  • Last commit: 2024-08-01 07:26:26 +0000
  • No local changes.
/home/gb001/research_bench/WritingBench
Research & benchmarks
Comprehensive benchmark for generative writing evaluation.
README.md; benchmark_query/; evaluator/; pics/; calculate_scores.py; evaluate_benchmark.py; generate_response.py; LICENSE; +1 more
Top languages: Other (8f/3,525l) Python (7f/677l) Markdown (1f/183l)
Benchmark/evaluation harness with scoring scripts and prompt assets.
255d · stale · git
16 files · 4,385 LOC · Small
git on main
  • Tracked: 18
  • Last commit: 2025-12-19 03:36:28 +0000
  • No local changes.
/home/gb001/src/llama.cpp
Tooling / memory systems
llama.cpp local LLM inference/build workspace.
README.md; pyproject.toml; requirements.txt; app/; benches/; build/; ci/; cmake/; +37 more
Top languages: Other (3512f/1,115,166l) CSV (11f/183,987l) Python (196f/59,963l) HTML (5f/35,214l)
Native C/C++ build artifacts, library, server, and inference CLIs.
87d · stale · git
4558 files · 1,488,557 LOC · Large
git on master
  • Tracked: 2948
  • Last commit: 2026-06-05 20:45:25 +0000
  • No local changes.
/home/gb001/openclaw-cli-ref
References / support
OpenClaw CLI reference/help notes rather than an active product repo.
help/
Top languages: Text (1f/1l)
Documentation/reference output only.
163d · stale · workspace
1 files · 1 LOC · Tiny
workspace
  • Non-git workspace.
  • Scanned files: 1
Source of truth: live filesystem scan + git metadata + curated project descriptions.
Generated 2026-08-31 10:36 UTC