codex usage

    Codex Daily Usage - July 15, 2026

    A production day shaped by two concentrated work blocks, delegated execution, and verified delivery across repositories.

    Period
    Jul 15, 2026 00:00 EDT to Jul 16, 2026 00:00 EDT
    Generated
    Jul 19, 2026
    Validation
    Passed
    Analyzer
    codex-usage-report-v4
    18
    Active rollouts

    11 top-level and 7 delegated.

    13
    Successful commit operations

    Verified from terminal outcomes; one operation amended an existing commit.

    125
    Files touched

    Unique paths in successful patch-application records.

    2
    Repositories delivered

    codex-skills and hyphenomenon

    Executive Summary

    July 15 was a delivery day, not a steady stream of tool activity. 91.2% of all tool calls occurred in two focused blocks: 3-6 AM and 11 AM-3 PM ET.

    The work used 18 active rollouts, including 7 delegated rollouts. hyphenomenon accounted for 74.6% of tool calls, while 504 tool calls (24.1%) happened outside resolved git metadata. Verified terminal outcomes show 13 successful commit operations and delivery output for 2 repositories.

    The Day Had Two Production Blocks

    Tool calls by local hour, with verified commit outcomes marked below. Labels remain visible without hover.

    12 AM
    2 AM
    319
    4c
    65
    1c
    4 AM
    239
    3c
    224
    2c
    6 AM
    5
    4
    8 AM
    49
    85
    10 AM
    251
    4
    12 PM
    391
    64
    2 PM
    348
    3c
    40
    4 PM
    6 PM
    8 PM
    10 PM

    Total: 2,088 tool calls. “c” marks successful commit operations in that hour.

    Where The Work Landed

    Tool volume and patch churn answer different questions, so they are shown side by side instead of collapsed into one repo count.

    Tool Calls By Repository

    Working repository resolved from tool workdir or turn context.

    hyphenomenon1,558 75%
    Outside git metadata504 24%
    bpe-marketing-lp15 1%
    codex-skills11 1%

    Patch Churn By Repository

    Added plus removed lines in patch inputs. Diagnostic only, not a productivity score.

    hyphenomenon4,681 52%
    Outside git metadata4,237 48%

    Delivery Extended Beyond One Repository

    hyphenomenon dominated resolved tool activity, but it was not the whole delivery footprint. The 504 tool calls outside git metadata represent 24.1% of the total and should not be mislabeled as hyphenomenon work. Terminal output names codex-skills and hyphenomenon as delivered repositories.

    Output Was Concentrated

    A small number of rollouts carried most patch churn, while file touches skewed toward documentation and governance work.

    Patch Churn By Rollout

    Rollout 122,910 33%

    Top-level | Outside git metadata

    Rollout 112,340 26%

    Top-level | hyphenomenon

    Rollout 011,577 18%

    Top-level | hyphenomenon

    Rollout 06783 9%

    Top-level | Outside git metadata

    Rollout 02544 6%

    Top-level | Outside git metadata

    Rollout 04269 3%

    Delegated | hyphenomenon

    Other495 6%

    File Touch Mix

    267touches
    Docs and governance17967%
    Application and pipeline code3613%
    Report prototypes228%
    Other files176%
    Config and automation135%

    Verified Delivery Ledger

    Outcomes come from terminal return records. Intent-only command counts remain in the technical appendix.

    Command Outcomes

    Commit operations - success
    Count13
    Share48%
    Content push operations - success
    Count10
    Share37%
    Branch cleanup pushes - success
    Count2
    Share7%
    Content push operations - pending
    Count1
    Share4%
    Pull request operations - unknown
    Count1
    Share4%

    Repositories In Successful Push Output

    codex-skills
    Count1
    Share50%
    hyphenomenon
    Count1
    Share50%

    Verification Was Part Of The Work

    The day produced 13 successful commit operations. One was an amend, so commit operations should not be read as 13 independent feature units. 1 content push ended in a long-running handoff rather than a terminal success record; later delivery evidence may still prove its commit reached the remote. Patch churn is useful for concentration and file-mix analysis, but it is not a productivity score.

    Chronicle Shows A Mixed Operating Day

    Chronicle supplies contextual coverage across desktop work. Lane counts overlap and should not be summed as duration.

    Work Lane Mentions

    Planning and knowledge work95 27%
    Software and repo delivery88 25%
    Storefront and catalog work68 19%
    Print and artwork production67 19%
    Wholesale and shipping36 10%

    Evidence Coverage

    Screenshot citations
    Count910
    Share71%
    Unique cited minutes
    Count265
    Share21%
    Overlapping 10-minute summaries
    Count105
    Share8%
    Observed clock hours
    Count10
    Share1%

    Chronicle Context

    Chronicle contributes 265 unique cited minutes across 10 observed clock hours. The summaries show a mixed operating day spanning planning, software delivery, storefront/catalog work, print production, and wholesale/shipping. These lane counts overlap because one ten-minute summary can mention several kinds of work; they are coverage signals, not additive duration.

    Technical telemetry appendix

    Raw tool, event, workspace, and patch diagnostics used to audit the analyzer. These counts are intentionally secondary to outcomes.

    Tool Surfaces

    Terminal/local shell1,742 83%
    Code editing178 9%
    Local image review50 2%
    Multi-agent tools31 1%
    Planning/status31 1%
    Long-running task waits16 1%
    Vercel15 1%
    Other25 1%

    Function Calls

    1,519
    95
    50
    31
    16
    13
    9
    49
    exec_command
    write_stdin
    view_image
    update_plan
    wait
    wait_agent
    js
    Other

    Custom Tool Calls

    apply_patch
    Count178
    Share58%
    exec
    Count128
    Share42%

    Response Events

    response_item
    Count6,488
    Share63%
    event_msg
    Count3,671
    Share36%
    turn_context
    Count87
    Share1%
    session_meta
    Count45
    Share0%
    world_state
    Count33
    Share0%
    compacted
    Count9
    Share0%
    inter_agent_communication_metadata
    Count4
    Share0%

    Patch Applications

    Apply patch calls
    Count178
    Share52%
    Successful patch applications
    Count166
    Share48%

    Output Records

    Tool output records
    Count2,086
    Share77%
    Git refs in output
    Count330
    Share12%
    Truncated outputs
    Count117
    Share4%
    Long-running command handoffs
    Count101
    Share4%
    Non-zero terminal exits
    Count70
    Share3%
    GitHub PR links emitted
    Count4
    Share0%

    Method Notes

    An active rollout is one JSONL rollout file with at least one timestamped record in the window. Top-level and delegated rollouts are reported separately. Chronicle summaries overlap by design. Patch lines count additions plus removals present in apply_patch inputs and include the work used to build this report, so they are shown as diagnostic context rather than a headline output metric. Message-role rates are excluded because system and assistant event volume does not represent human effort.