fix: read internal token from disk per-request, surface auth errors#97
Merged
Conversation
Two bugs: 1. Server: _verify_internal_auth cached the token in app.state at startup. If the token file was overwritten later (auto-start in another process), server and CLI disagreed -> 401 -> hermes node list silently showed all nodes as disconnected. Now reads ~/.hermes/nodes-internal-token on every request via new _read_token_from_disk() helper. Uses hmac.compare_digest. 2. CLI/tools: _connected_names and _node_list_impl silently swallowed auth errors. _connected_names now returns set[str] | None; None means "cannot query". 401/403 print a clear stderr warning. _cmd_list shows a warning banner and marks rows as "unknown" instead of "disconnected". New state constant: STATE_UNKNOWN. Signed-off-by: Blasius Patrick <blasius.patrick@gmail.com>
blaspat
force-pushed
the
fix/token-disk-read-auth-errors
branch
from
July 21, 2026 04:38
fc062de to
d9b3a64
Compare
blaspat
added a commit
that referenced
this pull request
Jul 21, 2026
* fix: read internal token from disk per-request, surface auth errors Two bugs: 1. Server: _verify_internal_auth cached the token in app.state at startup. If the token file was overwritten later (auto-start in another process), server and CLI disagreed -> 401 -> hermes node list silently showed all nodes as disconnected. Now reads ~/.hermes/nodes-internal-token on every request via new _read_token_from_disk() helper. Uses hmac.compare_digest. 2. CLI/tools: _connected_names and _node_list_impl silently swallowed auth errors. _connected_names now returns set[str] | None; None means "cannot query". 401/403 print a clear stderr warning. _cmd_list shows a warning banner and marks rows as "unknown" instead of "disconnected". New state constant: STATE_UNKNOWN. Signed-off-by: Blasius Patrick <blasius.patrick@gmail.com> * fix: rotating log for auto-start, quiet stdout - Replace print()+append with RotatingFileHandler (5 MB cap, 1 backup) - Only print lifecycle events to stdout (server started, failures, errors) - Routine diagnostics go to file only (port checks, thread lifecycle, etc.) - Remove stale import urllib.error (missed in #97) Signed-off-by: Blasius Patrick <blasius.patrick@gmail.com> --------- Signed-off-by: Blasius Patrick <blasius.patrick@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Two bugs that caused
hermes node listto silently show all nodes as "disconnected":Bug 1: Internal token cached at startup, never re-read
_verify_internal_authcached the token inapp.state.internal_tokenat gateway startup. If the token file was overwritten later (e.g. by auto-start in another profile), the server and CLI would disagree on the token → 401 →hermes node listreturned empty connected set → all nodes showed "disconnected."Fix:
_verify_internal_authnow reads~/.hermes/nodes-internal-tokenon every request via new_read_token_from_disk()helper. Comparison useshmac.compare_digest.Bug 2: CLI/tools silently swallowed auth errors
_connected_names()caught all exceptions and returned an empty set. A 401 from token mismatch looked identical to "no nodes connected."Fix:
_connected_namesreturnsset[str] | None—Nonemeans "cannot query server." 401/403 errors print a clear warning to stderr._cmd_listshows a warning banner and uses newSTATE_UNKNOWN = "unknown"instead of "disconnected" when the server is unreachable._node_list_impldetects auth-error responses and returns anerrorfield.