߲jK "UdZddlZddlZddlmZddlmZmZmZddl m Z m Z m Z ddl mZmZdZdZed d d d d dddd Zeeefed<dZdddddddddd Zeeefed<dZd ed!eed"efd#Zd$ed"eefd%Zd&ed'ed(ed"efd)ZGd*d+Zy),aFile-mutation verification footers and turn-completion explanations for ``AIAgent``. The footer tells the model (and user) when a claimed file mutation did not land; the explainer summarises why a turn ended without a final answer. Every method resolves through ``AIAgent``'s MRO. N)suppress)AnyDictOptional)_extract_error_preview_extract_file_mutation_targets#_extract_landed_file_mutation_paths)FILE_MUTATING_TOOL_NAMESfile_mutation_result_landedu⚠️ No reply: zy{model} didn't produce a reply this time, even after retries. Send `continue` to try again, or switch models with /model.z^the model provider didn't answer after all retries. Send /retry, or switch models with /model.zsstreaming stopped early and only a partial response was recovered. Send `continue` to resume from where it stopped.zano new content was produced this turn; showing recovered prior context. Send `continue` to retry.zthe request was cancelled by a new correction on every attempt, so the turn stopped instead of retrying forever. Your last correction is queued as the next message.zevery provider in the fallback chain kept failing over, so the turn stopped instead of retrying forever. Send `continue` or switch provider.zfthe per-turn iteration/cost budget was exhausted before a final answer. Send `continue` to keep going.zjthe local model's context window was too small to finish. Increase the context size or use a larger model.zthe turn stopped while a tool result was still pending and the model produced no follow-up text. Send `continue` to let it summarize.) empty_response_exhausted!all_retries_exhausted_no_responsepartial_stream_recoveryfallback_prior_turn_contentredirect_restart_limit_exceededrebuilt_restart_limit_exceededbudget_exhausted ollama_runtime_context_too_smallpending_tool_result_EXIT_REASON_EXPLANATIONS))interrupted_during_api_callz[the request was interrupted mid-call before a reply was received. Send `continue` to retry.)max_iterations_reachedz}the maximum tool-iteration limit was reached before a final answer. Send `continue` to keep going, or raise `max_iterations`.)error_near_max_iterationsztan error occurred near the iteration limit before a final answer. Check the tool output above, then send `continue`.)repeated_outer_errorszthe turn kept failing with repeated errors and was stopped early instead of retrying forever. Check the errors above, then send `continue` to retry.uthe turn was stopped because another process was compressing this session. Your message should already be saved — please send it again after compression completes.uthe turn was stopped because this session was rotated by context compression and its live continuation could not be adopted. The storage itself is healthy — refresh the client (or start a new turn) so it picks up the new session id, then send your message again.uthe turn was stopped because another Hermes process took over this session. Your reply was not saved — wait for the other process to finish, then send your message again.uthe turn was stopped because session storage was busy (another Hermes process was writing to the state database). Your message should already be saved — please send it again in a moment.uthe session database file was replaced while Hermes was running, so this message was not saved (a copy is kept in {home}/sessions/). Stop Hermes (`hermes {profile_arg}gateway stop`), run `hermes {profile_arg}doctor` — not `hermes {profile_arg}doctor --fix`, which would repair the wrong file in place — then start it again and send your message once more. Advanced recovery steps are in the log.u8another Hermes process still holds an old copy of the session database's write-ahead log, so Hermes stopped writing to keep the file safe and this message was not saved (a copy is kept in {home}/sessions/). Nothing is lost. Quit every Hermes process on this profile (Desktop app, `hermes {profile_arg}gateway stop`, dashboard, cron), run `hermes {profile_arg}doctor` — it names any process still holding the log — then start Hermes again and send your message once more. Do not run `doctor --fix` or delete any state.db files while they run. Guide: {recovery_docs}uthe turn was stopped because the state database reported structural corruption (the transcript would have been lost on restart). Freeing disk space will not help. Recovery options: 1. Run `hermes {profile_arg}doctor --fix` 2. Stop the gateway, then recover with: hermes {profile_arg}sessions recover --source {db_path} --inspect-only (if it reports recoverable) hermes {profile_arg}sessions recover --source {db_path} --output recovered-state.db — recovery snapshots the damaged file first; do NOT run `sqlite3 ... ".recover"` against the live state.db, a vulnerable sqlite3 CLI can corrupt it further 3. Restore from a backup in {backups_dir}/ Then send your message again.aTthe turn was stopped because the session search index (FTS5) is corrupt and could not be detached, so this message was not saved. The message store itself is not damaged: do not run recovery tools or restore a backup. Run `hermes {profile_arg}doctor --fix` (or restart Hermes, which repairs the index on open), then send your message again.zHermes couldn't save this conversation to disk, so it stopped rather than lose your messages. The disk is probably full: free some space (or fix the permissions on {home}/state.db), then send your message again.) compressioncompression_closed turn_leaselockedreplaced deleted_walcorrupt fts_indexdisk_PERSISTENCE_CAUSE_EXPLANATIONSa Hermes couldn't save this conversation, so it stopped rather than lose your messages. Possible causes: the drive is out of room, or another Hermes process is holding the database. Close other Hermes windows, run `hermes {profile_arg}doctor` to check storage, then send your message again.pathtask_idreturncJ ddlm}t|||xsd}tj jtj j|S#t$r?tj j tj j|}YwxYw)aOne key per on-disk target: the file tools' task-resolved absolute path, case-folded on case-insensitive hosts. A failure recorded as ``notes.md`` and the write that later lands as ``/repo/notes.md`` (or ``Notes.md`` on Windows) must meet on the same key.r)_resolve_path_for_taskdefault) tools.file_tools_pathsr(str Exceptionosr$abspath expandusernormcasenormpath)r$r%r(resolveds :/home/ubuntu/.hermes/hermes-agent/agent/turn_explainers.py_file_mutation_identityr4sy=A-dG4HyIJ 77  BGG,,X6 77 =77??277#5#5d#;<=sAAB"!B"identityc| tj|}|j|jfS#t$rYywxYw)zX``(mtime_ns, size)`` of the target, ``None`` when it does not exist (or cannot be read).N)r-statOSError st_mtime_nsst_size)r5sts r3_file_stat_signaturer<s> WWX  NNBJJ '' s / ;;env_var config_key cache_attrc tjj|}| |jj dvSt ||d}||S ddlm}|xsi}t|tr|jdnd}t|tr||vrt|j|} nd} t||| | S#t$ri}YpwxYw#t$rYywxYw)ub``display.`` (default True), cached per agent on ``cache_attr``. ``env_var`` overrides on every call and is never cached. Reads the persisted config.yaml so gateway and CLI share the setting; ``load_config`` is imported lazily (startup cycle, and tests patch it at ``hermes_cli.config``). Any failure → True (safe default: on).N>0noofffalser) load_configdisplayT) r-environgetstriplowergetattrhermes_cli.configrEr, isinstancedictboolsetattr) agentr=r>r?envcached _load_config_cfg_displayenableds r3_display_flag_enabledrXs jjnnW% ?99;$$&.II I D1  M  E>'RD+5T4*@488I&d h %**@8<< 34GGz7+ D  s=ACCC &A"C CCCC C&%C&c 8eZdZdZdddedeeefdededeed df d Z e d eeeeeffd eeeeefffd Z d efd Z d efdZ ejdZeded efdZed eeeeeffd efdZe ddedeeded efdZy)TurnExplainersMixinzPFile-mutation failure footer + turn-completion explainer (see module docstring).N)r% tool_nameargsresultis_errorr%r&c  |tvryt|dd}|yt||}|syt||}|rt |||} t|dd} | | j | t|dd} | Ot| ddrBddlm} | |xsd 0| D]+} tt5| j| ddd-|rA|s?t|}|D].}t||}|j||||t|d 0y|r n|Dchc]}t||}}t|j!D]6\}}|j#d t|||vs%|j%|d8y#1swYxYwcc}w) aRecord a ``write_file`` / ``patch`` outcome for the turn-end verifier. Failures store ``{path: {error_preview, tool, identity, stat}}`` keyed by the model's spelling; ``identity`` is the resolved on-disk target and ``stat`` its signature at failure time. A later success on the same identity (any spelling) removes the entry. No-op when the per-turn state dict is not initialised (tool dispatched outside ``run_conversation``). N_turn_failed_file_mutations_turn_file_mutation_paths_checkpoint_mgrrWFr)container_backend_for_taskr))tool error_previewr5r7r5)_FILE_MUTATING_TOOLSrKrr r updater*rcrr,record_agent_writerr4 setdefaultr<listitemsrHpop)selfr[r\r]r^r%statetargetslanded landed_pathschangedmgrrc_ppreviewr$r5pclearedinfos r3_record_file_mutation_resultz0TurnExplainersMixin._record_file_mutation_results 0 0 ;TB = 0DA ,Y? >y$PVWLd$?FG"|,$ 148C73 5#AM-g.BCK*7%i072226777 F,V4G 24A  % (2Fx2P( +1,g(73G#5;;=1 * d88J(?g(NOSZZIIdD) *!77s&E5 F5E> failedc|jDcic] \}}d|vst|d|dk(r||"c}}Scc}}w)aDrop entries whose target changed on disk since the failed call. The recorder only sees write_file/patch receipts; a terminal redirect or an execute_code write leaves none. Re-checking the stat signature at turn end keeps the footer from listing a file that was in fact modified later this turn. Entries without a snapshot (hand-built dicts) are kept as-is. r7r5)rkr<)rzr$rxs r3_file_mutations_still_failedz0TurnExplainersMixin._file_mutations_still_failedsQ*0 %4T!%9$z:J%KtTZ|%[ $J   s%=c t|dddS)zZ``display.file_mutation_verifier`` / ``HERMES_FILE_MUTATION_VERIFIER`` (a patchable seam).HERMES_FILE_MUTATION_VERIFIERfile_mutation_verifier%_file_mutation_verifier_enabled_cacher=r>r?rXrms r3_file_mutation_verifier_enabledz3TurnExplainersMixin._file_mutation_verifier_enabled(s$ 9F^>  c t|dddS)zM``display.turn_completion_explainer`` / ``HERMES_TURN_COMPLETION_EXPLAINER``. HERMES_TURN_COMPLETION_EXPLAINERturn_completion_explainer(_turn_completion_explainer_enabled_cacherrrs r3"_turn_completion_explainer_enabledz6TurnExplainersMixin._turn_completion_explainer_enabled/s$ z>TurnExplainersMixin._neutralize_footer_paths..Ds1QWWQZL1Br)_FOOTER_PATH_REsub)clsrs r3_neutralize_footer_pathsz,TurnExplainersMixin._neutralize_footer_paths<s& K""&&'BDIIrc |sydt|dg}t|jdd}|D][\}}|jdxsdj }|jdxsd}|j d |d |d |xsd ]t|t|z }|d kDr|j d|d|j dj|S)a/Render the per-turn failed-mutation dict as a user-facing footer. Up to 10 paths with their first error preview, then an overflow count; "" when nothing failed. Every path is backtick-wrapped via ``_neutralize_footer_paths`` so protected files cannot be auto-delivered. u⚠️ File-mutation verifier: z file edit(s) FAILED this turn despite any wording above that may suggest otherwise. Run `git status` or `read_file` to confirm what actually landed.N rerdpatchu • `u` — [z] rzru • … and z more )lenrjrkrHrIappendrjoin) rrzlinesshownr$rxrurd remainings r3$_format_file_mutation_failure_footerz8TurnExplainersMixin._format_file_mutation_failure_footerFs -6{m; ;  V\\^$Sb) OJD$xx06B==?G88F#.wD LL74&vR8K87LM N OK#e*, q= LL>)E: ;++DIIe,<==rturn_exit_reasonpersistence_causemodelc|syt|}|jdrytj|}|#tD]\}}|j|s|}n|d|vr|j |xsd}||dk(rddlm}m} dd l m } tj|xsd tjd |jd | jd | }|dvrSddlm} ddlm} |jdt|xs| }|jdt| dz }|r t"|zSdS)aUser-facing explanation for an abnormal turn ending, or "" for normal / unknown reasons. ``text_response(...)`` is the healthy terminal; unknown/diagnostic-only reasons (e.g. ``guardrail_halt``, which surfaces its own message) are not second-guessed. r text_responsez{model}z The model)rsession_persistence_failedr)display_hermes_homeprofile_cli_selector)STORAGE_RECOVERY_DOCS_URLunknownz{home}z {profile_arg}z{recovery_docs})r r!)get_default_hermes_root)_default_db_pathz {db_path}z {backups_dir}backups)r+ startswithrrH _EXIT_REASON_PREFIX_EXPLANATIONSformathermes_constantsrrhermes_state_errorsrr# _PERSISTENCE_DEFAULT_EXPLANATIONreplacer hermes_stater _NO_REPLY) rrdb_pathrreasonbodyprefixrrrrrrs r3#_format_turn_completion_explanationz7TurnExplainersMixin._format_turn_completion_explanationasO %&   _ -(,,V4 < @  $$V,D    T 1;;U%9k;:D *@A*,EF !$<<D9||KW5R@P@R1ST||#S)@)BY)N%O$(y4/R/r)NNr)__name__ __module__ __qualname____doc__r+rrrOrry staticmethodr|rrrecompiler classmethodrrrrr3rZrZs]Z%)2*2*$(cN2*$sDcN?R:S>X[>>4ce,0,02:3-,0]`,0 ,0,0rrZ)rr-r contextlibrtypingrrragent.tool_dispatch_helpersrrr agent.tool_result_classificationr rfr rEMPTY_RESPONSE_EXPLANATIONrr+__annotations__rr#rr4tupler<rOrXrZrrr3rsB  && B!; 5 F 3 4    7 ; C%-4S>%P$ 2 F  4   ,  D (& # =QL3c3hL\-! 8# 8 8# 8(3(8E?(ScsW[:k0k0r