Skip to content

docs: correctness sweep — 15 verified fixes (Bucket 1) - #34781

Merged
teknium1 merged 18 commits into
mainfrom
hermes/hermes-672c8746
May 29, 2026
Merged

docs: correctness sweep — 15 verified fixes (Bucket 1)#34781
teknium1 merged 18 commits into
mainfrom
hermes/hermes-672c8746

Conversation

@teknium1

Copy link
Copy Markdown
Contributor

Summary

Fixes 15 docs-correctness defects on the Docusaurus site — invalid commands, wrong defaults, broken images/links, and stale config shapes — each verified against current main before inclusion.

This is the "Bucket 1" correctness sweep: 39 candidate PRs were checked claim-by-claim against the codebase; 15 were confirmed real and salvaged here, the rest dropped (false/stale claims) or held (bundled feature-code).

Changes

Validation

Result
Every correctness claim verified against current main (commands, config keys, env-var units, link targets, runtime ordering)
Docs build (npm run build) passes — EN + zh-Hans both succeed
New broken links/anchors zero on English pages (remaining warnings are pre-existing zh-Hans translation lag)
Touched SVG images confirmed resolve in the built output
Authorship preserved per-commit via rebase-merge; two fabricated-identity commits re-attributed to real GitHub authors (@Twanislas, and a Matrix PR held out)

Dropped during verification (false/stale): hermes login refs (command exists), "hallucinated backends" (all real), matrix !-prefix (dispatch is /), already-fixed TTS table + broken-link batch. Held for separate review: honcho cloud-disclosure (plugin code) and a Matrix device-verification script (362 LOC + tests).

Infographic

docs-correctness-sweep

Hashclaw and others added 18 commits May 29, 2026 11:29
- ACL LaTeX template comment: seperate -> separate
- CONTRIBUTING and docs site: builtin trust -> built-in trust (prose/table cells)

Made-with: Cursor
hermes-ai/hermes-agent → NousResearch/hermes-agent (2 occurrences).
The old org name leads to 404 pages.
The Google Workspace skill doc linked to https://github.com/nicholasgasior/gws
which returns 404. The actual upstream CLI lives at
https://github.com/googleworkspace/cli (the official Google Workspace CLI in
Rust, dynamically built from the Google Discovery Service).

Closes #28922
Fix download link for Linux/macOS binary in documentation.
- Replace outdated linear ordering in prompt-assembly guide with
  current stable/context/volatile tier contract from system_prompt.py
- Clarify where memory/profile snapshots live versus skills guidance
- Document that pre_llm_call context is user-message injection, not
  cached system-prompt mutation
- Update architecture guide wording to reference system_prompt.py +
  prompt_builder.py tiered assembly

Closes #34118
Use the current top-level fallback_providers list in fallback docs and keep fallback_model documented only as the legacy compatibility shape. Also align cron and delegation fallback coverage with current runtime behavior.

Closes #19691

Co-authored-by: Codex <codex@openai.com>
…pages

Fixes #24809

The docs site uses baseUrl='/docs/' but the <img> tags in sessions.md
and cli.md referenced images at /img/docs/... which resolves to a 404.
The static files are served at /docs/img/docs/... instead.

Before: <img src="/img/docs/session-recap.svg"> → 404
After:  <img src="/docs/img/docs/session-recap.svg"> → 200

Also fixes cli-layout.svg which had the same issue.
…how'

'hermes config get <key>' is referenced in three guides but is not a
valid subcommand. The valid subcommands under 'hermes config' are
{show,edit,set,path,env-path,check,migrate}. 'hermes config show' is
already used elsewhere in the docs (including 'hermes config show |
grep <pattern>' in the FAQ), so it's the idiomatic replacement.

- work-with-skills.md: 'View all skill config' now uses
  'hermes config show | grep ^skills\.config'
- migrate-from-openclaw.md: session-policy check now reads the value
  from 'hermes config show'
- configuring-models.md: 'inspect what the CLI will actually use'
  now uses 'hermes config show | grep ^model\.'

Refs #30195
…e semantics

The curator docs stated that any skill not bundled/hub-installed was
'agent-created' and subject to curation — including foreground-created
skills and hand-written ones. Since PR #19621 (May 2026), the curator
requires an explicit  marker in .usage.json, which
only the background self-improvement review fork sets.

Changes:
- Rewrite 'What agent-created means' to document the 3-step eligibility
  check (not bundled + not hub + created_by=agent marker)
- Explain that foreground skill_manage(create) does NOT mark skills as
  agent-created (user-directed by design)
- Warn that hand-written skills are NOT curated
- Add note in Per-run reports explaining the '(not resolved)' display
  when no candidates exist (LLM pass skipped, not a config error)
- Link to skill_provenance.py for the write-origin ContextVar

Ref: PR #19621, tools/skill_provenance.py, tools/skill_manager_tool.py
@github-actions

Copy link
Copy Markdown
Contributor

🔎 Lint report: hermes/hermes-672c8746 vs origin/main

ruff

Total: 0 on HEAD, 0 on base (➖ 0)

🆕 New issues: none

✅ Fixed issues: none

Unchanged: 0 pre-existing issues carried over.

ty (type checker)

Total: 9454 on HEAD, 9454 on base (➖ 0)

🆕 New issues: none

✅ Fixed issues: none

Unchanged: 4908 pre-existing issues carried over.

Diagnostics are surfaced as warnings — this check never fails the build.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

P3 Low — cosmetic, nice to have type/docs Documentation improvements

Projects

None yet

Development

Successfully merging this pull request may close these issues.