# Monitoring and Logging Kit: Apache projects for small business

October 29, 2026. [Read the article](https://jwatte.com/blog/apache-monitoring-logging-small-business/).

## How to use this kit

Choose one service whose failure interrupts real work. Write the response action and the responsible person before configuring alerts. Use the cards to distinguish a service check from application tracing and log collection. These are researched starting points, not tested installations; no avoided-downtime savings are assumed.

## HertzBeat starter card

HertzBeat checks services, applications, databases and websites and sends notifications. It fits a capable owner or helper who understands Docker, service access and restores. Example uses: a twelve person auto repair business with two locations watches its booking systems; a three-site car wash with six staff watches service health and certificate expiry. Both assume three existing watched hosts and one separate monitoring server.

The [quick start](https://hertzbeat.apache.org/docs/start/quickstart/) gives two CPU cores, 4 GB RAM and 10 GB disk, with Docker 20.10 or newer or Java 25 or newer. Those are base requirements, not a capacity promise for all database and retention loads. The [Compose route](https://hertzbeat.apache.org/docs/start/docker-compose-deploy/) recommends HertzBeat, PostgreSQL and VictoriaMetrics. Prefer that documented stack; the [minimal single-container page](https://hertzbeat.apache.org/docs/start/docker-deploy/) says its setup is not recommended for production. Start with the [1.9.0 download](https://hertzbeat.apache.org/docs/download/), released September 17, 2026, and pin the reviewed version.

Backup and update checklist:

- Preserve configuration, persistent database volumes and the secrets required to restore access.
- Review [the security model](https://hertzbeat.apache.org/docs/help/security_model/) and limit who can configure templates, URLs and plugins.
- Keep a record of check destinations, notification routes and retention settings.
- Restore a copy away from the working system, then test notifications without paging staff unnecessarily.
- Rehearse updates against the copy and retain a rollback procedure.

First week: day one choose a service and response owner; day two install the documented stack and restrict access. Day three add a single useful check. Day four trigger an agreed controlled failure and recovery. Day five restore a backup and let the responder follow the written instructions. Keep the monitor outside the failure it is meant to detect; add more checks only when their messages lead to useful action.

## Provider cards

**SkyWalking.** Application agents, storage, trace fields and retention require IT. [Backend setup](https://skywalking.apache.org/docs/main/v11.0.0/en/setup/backend/backend-setup/) requires Java 11 at minimum and lists Java 11, 17, 21 and 25; [Docker instructions](https://skywalking.apache.org/docs/main/v11.0.0/en/setup/backend/backend-docker/) document a one-host Compose route with storage, backend and UI. No universal numeric hardware minimum is published there. Our 8 GiB one-host evaluation size has no failover and must pass workload testing. [Downloads](https://skywalking.apache.org/downloads/) lists core 11.0.0, August 28, 2026; agents have separate versions. Ask which fields leave the application, how collection is disabled and who tunes retention. Default deployment must be restricted to operations under the [security guidance](https://skywalking.apache.org/docs/main/v11.0.0/en/security/readme/). Datadog infrastructure plus APM is a managed alternative; instrumentation is separately quoted on both sides. Allow 1 to 3 monthly administration hours for our modest single-host pilot, not a cluster or an on-call contract.

**Flume.** This is an inherited-system card, not a recommendation for a new installation. [Downloads](https://flume.apache.org/download.html) lists 1.11.0; the [homepage](https://flume.apache.org/) dates it October 24, 2022 and [Logging](https://logging.apache.org/) now lists Flume inactive. The [guide](https://flume.apache.org/FlumeUserGuide.html) documents one JVM agent, Java 8 or newer and configuration-dependent memory/disk needs; no universal numeric hardware floor is given. Our 4 GiB agent server is a planning allowance and excludes its destination. Allow 1 to 3 monthly administration hours while preparing replacement. Ask the provider to trace a known event across restart and destination failure, preserve configuration and recovery state and explain loss/duplication behavior. [Grafana Cloud](https://grafana.com/pricing/) is a possible managed destination with its own collector and migration work; it is not hosted Flume.

**OzHera (Incubating).** The [deployment guide](https://ozhera.incubator.apache.org/en/docs/deployment.html) assumes Kubernetes operations knowledge and dependencies including database, messaging, search and monitoring services. No universal CPU/RAM/node minimum is published. The smallest verified architecture is an existing Kubernetes environment with persistent storage and documented dependencies. Ask a provider to size it, review authentication defaults and quote integration. Our initial assumption is 24 to 60 setup hours and 6 to 16 monthly administration hours, plus actual hosting and existing-platform allocation. [GitHub](https://github.com/apache/ozhera/releases/tag/v2.2.6-incubating) confirms 2.2.6-incubating, June 17, 2026; the website download list is older. Datadog infrastructure plus APM or Grafana Pro are competing managed services with published base prices below, not official OzHera support.

## Complete group table

| Project | What it does | Who runs it | Latest release and date |
|---|---|---|---|
| [Apache Chainsaw](https://logging.apache.org/) | Helps developers inspect and filter application logs in a desktop viewer. | Your developer | [2.1.0; 2021-06-12](https://downloads.apache.org/logging/chainsaw/2.1.0/) |
| [Apache Commons Logging](http://commons.apache.org/logging/) | Gives Java libraries a common interface to different logging systems. | Your developer | [1.3.5; 2025-02-08](https://projects.apache.org/json/foundation/projects.json) |
| [Apache Flume](http://flume.apache.org) | Collects logs and moves them into shared storage for analysis. | Your IT department or provider | [1.11.0; 2022-10-24](https://flume.apache.org/) |
| [Apache HertzBeat](https://hertzbeat.apache.org) | Monitors servers and services and alerts staff when something needs attention. | You | [1.9.0; 2026-09-17](https://hertzbeat.apache.org/docs/download/) |
| [Apache log4cxx](https://logging.apache.org/log4cxx) | Records application events and errors in C++ software. | Your developer | [1.8.0; 2026-07-28](https://downloads.apache.org/logging//log4cxx/1.8.0/) |
| [Apache Log4j](https://logging.apache.org/log4j/2.x/index.html) | Records application events and errors in Java software. | Your developer | [2.26.1; 2026-07-02](https://github.com/apache/logging-log4j2/releases/tag/rel/2.26.1) |
| [Apache log4net](https://logging.apache.org/log4net/) | Records application events and errors in .NET software. | Your developer | [3.5.0; 2026-09-25](https://github.com/apache/logging-log4net/releases/tag/rel/3.5.0) |
| [Apache log4php](https://logging.apache.org/) | Records application events and errors in PHP software. | Your developer | [2.3.0; 2013-01-31](https://downloads.apache.org/logging/log4php/2.3.0/) |
| [Apache Otava (Incubating)](https://otava.incubator.apache.org/) | Detects statistically significant changes in recorded performance measurements. | Your developer | [0.8.0-incubating; 2026-05-08](https://github.com/apache/otava/releases/tag/0.8.0-incubating) |
| [Apache OzHera (Incubating)](https://ozhera.incubator.apache.org/) | Combines application metrics, traces, logs and alerts for operations teams. | Your IT department or provider | [2.2.6-incubating; 2026-06-17](https://github.com/apache/ozhera/releases/tag/v2.2.6-incubating) |
| [Apache SkyWalking](http://skywalking.apache.org) | Tracks application performance, service calls, metrics and logs. | Your IT department or provider | [11.0.0; 2026-08-28](https://skywalking.apache.org/downloads/) |

Chainsaw, Flume and log4php are inactive according to the Logging parent site. Chainsaw/log4cxx/log4php dates are artifact modification dates; Commons Logging's date is registry evidence. SkyWalking's version is the core server version. Otava and OzHera are not yet fully endorsed by the ASF while incubating; see the [Incubator guidance](https://incubator.apache.org/guides/branding.html). Developer libraries still require patch ownership inside the applications that contain them.

## Cost worksheet

Fill low/high values: observation-server count ___ / ___; server price ___ / ___; watched hosts ___ / ___; setup hours ___ / ___; monthly admin hours ___ / ___; backup GB ___ / ___; monthly ingested logs ___ / ___; retention ___ / ___. Record which destination and interface are included. Add collector migration, instrumentation, usage overages and response contracts as separate quoted lines.

Our assumptions use the series research read October 29, 2026: $68.09 per hour for administration, derived from the [BLS administrator wage](https://data.bls.gov/oesprofile/?areas=INDUSTRY,STATE,MSA&major_group=150000&measure=01&occupation=151244&year=2025) of $47.66 divided by the [70.0% wages share of compensation](https://data.bls.gov/timeseries/CMU2020000000000P). This estimates employee compensation. An outside provider bills more; get a quote.

[DigitalOcean monthly prices](https://www.digitalocean.com/pricing/droplets) are $24 for 2 vCPU and 4 GiB RAM, or $48 for 4 vCPU and 8 GiB RAM. [Backblaze B2](https://www.backblaze.com/cloud-storage/pricing) is $6.95 per decimal TB per month, with the first 10 GB free. Our low and high backup assumptions are 50 and 100 GB retained, including versions: (50 minus 10) × $0.00695 = $0.278, or (100 minus 10) × $0.00695 = $0.6255 monthly. We assume no chargeable excess transfer or API calls. Check both providers' allowances before moving large files.

For one server, our low and high labor assumptions are 4 and 12 setup hours, then 1 and 3 administration hours per month. For a small cluster they are 24 and 60 setup hours, then 6 and 16 hours per month. These are planning ranges, not measured installation times. Custom development, data cleanup, migration, training, compliance reviews and major incidents are excluded. Budget those separately on both sides.

Setup = hours × $68.09. Monthly = server charges + backups + administration hours × $68.09. The 36 month total = setup + 36 × monthly. Savings = alternative total minus self hosted total. Break even = setup difference divided by monthly savings, rounded up, only when monthly savings are positive. Displayed totals are rounded to whole dollars; calculations retain cents. Prices are US dollars before tax, read October 29, 2026.

Each self hosted worked case has one observation or collection server at both ends, 50 to 100 GB retained backups, 4 to 12 setup hours and 1 to 3 monthly admin hours. Existing application-server costs are excluded from both alternatives. SaaS incremental setup is assumed zero; custom collection, instrumentation and migration require quotes on both sides. These are operating budgets, not equivalent feature certifications or savings from hypothetical prevented incidents.

### HertzBeat and three-host infrastructure monitoring

One $24 server. [Datadog Infrastructure Pro](https://www.datadoghq.com/pricing/), read October 29, 2026, is $15/host/month annually billed or $18 on demand. We choose three fixed hosts on demand: 3 × $18 = $54 monthly. Website synthetic and certificate features need separate verification; do not assume every check is included in Infrastructure Pro.

| Cost over 36 months | Low assumption | High assumption |
|---|---:|---:|
| Self hosted setup | $272 | $817 |
| Self hosted monthly | $92 | $229 |
| Self hosted 36 months | $3,598 | $9,057 |
| Datadog Infrastructure Pro, three hosts 36 months | $1,944 | $1,944 |
| Self hosted savings | Negative $1,654 | Negative $7,113 |
| Break even month | never | never |

Low formula: 4 × $68.09 + 36 × ($24 + $0.278 + $68.09). High: 12 × $68.09 + 36 × ($24 + $0.6255 + 3 × $68.09). Alternative: 36 × 3 × $18. No break even.

### SkyWalking and two instrumented hosts

One $48 evaluation server. [Datadog APM](https://www.datadoghq.com/pricing/?product=application-performance-monitoring), read October 29, 2026, is $31/host/month annually billed or $36 on demand, with infrastructure monitoring attached. Adding Infrastructure Pro gives $46 annual-equivalent or $54 on demand per host. Our fixed two-host case uses 2 × ($36 + $18) = $108 monthly, before additional usage.

| Cost over 36 months | Low assumption | High assumption |
|---|---:|---:|
| Self hosted setup | $272 | $817 |
| Self hosted monthly | $116 | $253 |
| Self hosted 36 months | $4,462 | $9,921 |
| Datadog infrastructure and APM, two hosts 36 months | $3,888 | $3,888 |
| Self hosted savings | Negative $574 | Negative $6,033 |
| Break even month | never | never |

Use the preceding self hosted formula with $48 replacing $24. Alternative: 36 × $108. No break even. Application instrumentation and any extra retention or ingestion need their own quote.

### An inherited Flume agent versus a managed destination

One $24 agent server, with destination costs excluded from the self hosted floor. [Grafana Cloud Pro](https://grafana.com/pricing/), read October 29, 2026, has a $19 monthly platform fee plus usage, with 50 GB monthly log ingestion and 30-day retention included. Our 10 to 40 GB monthly log-volume assumption and all other service use stay inside included allowances.

| Cost over 36 months | Low assumption | High assumption |
|---|---:|---:|
| Self hosted setup | $272 | $817 |
| Self hosted monthly | $92 | $229 |
| Self hosted 36 months | $3,598 | $9,057 |
| Grafana Cloud Pro within allowances 36 months | $684 | $684 |
| Self hosted savings | Negative $2,914 | Negative $8,373 |
| Break even month | never | never |

Self hosted formulas match HertzBeat's dedicated server. Alternative: 36 × $19. No break even in this base comparison. Grafana includes a destination and interface while the Flume floor does not, so this is a prompt to obtain a replacement quote, not interchangeable service pricing. Collector and migration work are excluded from both sides and must be quoted.

### Setup time when Claude or ChatGPT does the typing

See the [post's AI setup comparison](https://jwatte.com/blog/apache-monitoring-logging-small-business/#setup-time-when-claude-or-chatgpt-does-the-typing) for the same HertzBeat example and tool roles. We keep one $24 monitoring server, three existing watched hosts and 50 to 100 GB retained backups, costing $0.278 to $0.6255 monthly. Hosting plus backup H is therefore $24.278 to $24.6255 monthly. Our paired low/high assumptions use setup shares q of 50%/80% and admin shares r of 75%/100%. The resulting hours include prompting, supervision, review, tests and corrections. They are planning assumptions, not measured deployment times.

| HertzBeat planning case | Low | High |
|---|---:|---:|
| Manual setup hours | 4 | 12 |
| AI assisted setup hours | 2 | 9.6 |
| Manual monthly admin hours | 1 | 3 |
| AI assisted monthly admin hours | 0.75 | 3 |
| Manual 36 month total | $3,598 | $9,057 |
| Claude Pro subscription, 36 months | $3,049 | $9,094 |
| ChatGPT Plus subscription, 36 months | $3,049 | $9,094 |
| API: Claude Opus 5.5, 36 months | $2,863 | $9,008 |
| API: GPT-6 Astra, 36 months | $2,886 | $9,179 |
| Datadog Infrastructure Pro, three hosts, 36 months | $1,944 | $1,944 |

Manual total = S × $68.09 + 36 × (H + A × $68.09). AI total = S × q × $68.09 + T0 + 36 × (H + A × r × $68.09 + Tm). S/A are manual setup/monthly hours; q/r are AI shares; H is hosting plus backup; T0/Tm are setup/monthly tool costs. Round only final totals to whole dollars.

Our assumed setup review of 1 to 3 hours and monthly review of 0.5 to 1.5 hours are already inside the AI hours. Do not add them again or bill unattended agent runtime as human work. Existing application-host costs, custom instrumentation, migration and major incidents remain excluded on both sides. Datadog's $1,944 comparison uses three hosts at $18 monthly on demand; feature and usage qualifications in the original worksheet still apply. AI assistance does not change the conclusion: the managed base is cheaper and self hosting never breaks even at this scope.

Claude Pro and ChatGPT Plus are $20 monthly, read October 29, 2026. Charge T0 = $20 for a separate setup month, followed by Tm = $5 for each of 36 operating months only if other work uses the remaining 75%. Total allocations across projects cannot exceed the subscription. With dedicated use, Tm = $20 adds $540 to either subscription total. Plan allowances vary; wait for a reset or price extra usage if necessary. These subscription amounts do not include API tokens or promise unlimited usage. We deduct no conditional API credits.

Our assumed cumulative setup budget is 200,000 to 1,000,000 input tokens and 20,000 to 100,000 output tokens. Monthly administration uses 100,000 to 500,000 input and 10,000 to 50,000 output. Repeated context, retries and billed reasoning are included. Low input is 30% ordinary input, 10% short cache writes and 60% cache reads. High input is entirely uncached. Cache writes replace ordinary input charges; categories never overlap. Use the applicable short context band for each request; monthly sessions may need fresh cache writes.

API tool cost = (ordinary input tokens × input rate + cache write tokens × write rate + cache read tokens × read rate + output tokens × output rate) / 1,000,000.

Rates per million tokens, in input/output/short write/read order, are $4/$20/$5/$0.20 for Claude Opus 5.5 and $10/$50/$12.50/$1 for GPT-6 Astra, at standard speed and global routing. Opus low setup is (60,000 × $4 + 20,000 × $5 + 120,000 × $0.20 + 20,000 × $20) / 1,000,000 = $0.764; monthly is $0.382. High setup/monthly charges are $6/$3. Astra's corresponding low charges are $1.97/$0.985 and high charges are $15/$7.50. Put these amounts into T0 and Tm separately from labor and infrastructure. No batch discount is assumed.

Our budget uses free local Desktop Commander or Codex CLI. Optional Desktop Commander Remote Pro costs $20 monthly; a setup month plus the same justified 25% ongoing allocation adds $200 over the modeled period. Free Remote allows 10,000 calls monthly; those are tool calls, not model tokens. Paid Remote is not required for the base comparison.

The high subscription case becomes more expensive than manual work in month 29 because its admin hours do not fall. AI can also take longer: testing 119% of manual hours, inspired by the METR 2025 result as a sensitivity rather than a prediction, adds about $717 to the low manual total after the shared subscription fees. Keep review time and do not infer deployment speed from a coding study.

Use a staging server, never paste secrets into chat, and approve commands or use a restricted sandbox. Watch provider and Cloudflare actions, and check HertzBeat alert destinations and permissions. Use supported supervised subscription tools, and API keys for custom automated pipelines; do not share personal account credentials. Review changes, test an alert and a restore, and retain responsibility for backups and updates.

## Acceptance and support

Trigger a controlled failure, confirm the alert and recovery message, and have the actual responder follow the instructions. Test log counts during a restart and destination outage. Remove secrets before collection and set retention. Check PCI DSS boundaries before collecting payment-related requests; the example excludes payment data and makes no compliance claim.

Read [ASF security guidance](https://www.apache.org/security/), [Flume advisories](https://flume.apache.org/security.html) and [Log4j security notices](https://logging.apache.org/log4j/2.x/security.html). Log4Shell fixes from 2021 are not a current application patch plan; ask vendors for supported releases and component inventories. [HertzBeat community help](https://hertzbeat.apache.org/docs/community/contact/) has no verified paid response guarantee in our research. Obtain a contract for required response times.

The [Apache License 2.0](https://www.apache.org/licenses/LICENSE-2.0) permits commercial use under its terms and distribution-notice requirements, with no warranty or trademark grant. This is not legal advice.

## Sources

1. [quick start](https://hertzbeat.apache.org/docs/start/quickstart/). Read October 29, 2026, 21:38 UTC.
2. [Compose route](https://hertzbeat.apache.org/docs/start/docker-compose-deploy/). Read October 29, 2026, 21:38 UTC.
3. [minimal single-container page](https://hertzbeat.apache.org/docs/start/docker-deploy/). Read October 29, 2026, 21:39 UTC.
4. [1.9.0 download](https://hertzbeat.apache.org/docs/download/). Read October 29, 2026, 21:37 UTC.
5. [the security model](https://hertzbeat.apache.org/docs/help/security_model/). Read October 29, 2026, 21:39 UTC.
6. [Backend setup](https://skywalking.apache.org/docs/main/v11.0.0/en/setup/backend/backend-setup/). Read October 29, 2026, 21:39 UTC.
7. [Docker instructions](https://skywalking.apache.org/docs/main/v11.0.0/en/setup/backend/backend-docker/). Read October 29, 2026, 21:39 UTC.
8. [Downloads](https://skywalking.apache.org/downloads/). Read October 29, 2026, 21:37 UTC.
9. [security guidance](https://skywalking.apache.org/docs/main/v11.0.0/en/security/readme/). Read October 29, 2026, 21:39 UTC.
10. [Downloads](https://flume.apache.org/download.html). Read October 29, 2026, 21:39 UTC.
11. [homepage](https://flume.apache.org/). Read October 29, 2026, 21:38 UTC.
12. [Logging](https://logging.apache.org/). Read October 29, 2026, 21:38 UTC.
13. [guide](https://flume.apache.org/FlumeUserGuide.html). Read October 29, 2026, 21:38 UTC.
14. [Grafana Cloud](https://grafana.com/pricing/). Read October 29, 2026, 21:39 UTC.
15. [deployment guide](https://ozhera.incubator.apache.org/en/docs/deployment.html). Read October 29, 2026, 21:38 UTC.
16. [ozhera release v2.2.6-incubating on GitHub](https://github.com/apache/ozhera/releases/tag/v2.2.6-incubating). Read October 29, 2026, 21:38 UTC.
17. [2.1.0; 2021-06-12](https://downloads.apache.org/logging/chainsaw/2.1.0/). Read October 29, 2026, 21:26 UTC.
18. [Apache Commons Logging](http://commons.apache.org/logging/). Read October 29, 2026, 21:13 UTC.
19. [1.3.5; 2025-02-08](https://projects.apache.org/json/foundation/projects.json). Read October 29, 2026, 21:00 UTC.
20. [Apache HertzBeat](https://hertzbeat.apache.org). Read October 29, 2026, 21:13 UTC.
21. [Apache log4cxx](https://logging.apache.org/log4cxx). Read October 29, 2026, 21:13 UTC.
22. [1.8.0; 2026-07-28](https://downloads.apache.org/logging//log4cxx/1.8.0/). Read October 29, 2026, 21:22 UTC.
23. [Apache Log4j](https://logging.apache.org/log4j/2.x/index.html). Read October 29, 2026, 21:13 UTC.
24. [logging-log4j2 release rel/2.26.1 on GitHub](https://github.com/apache/logging-log4j2/releases/tag/rel/2.26.1). Read October 29, 2026, 21:15 UTC.
25. [Apache log4net](https://logging.apache.org/log4net/). Read October 29, 2026, 21:13 UTC.
26. [logging-log4net release rel/3.5.0 on GitHub](https://github.com/apache/logging-log4net/releases/tag/rel/3.5.0). Read October 29, 2026, 21:15 UTC.
27. [2.3.0; 2013-01-31](https://downloads.apache.org/logging/log4php/2.3.0/). Read October 29, 2026, 21:26 UTC.
28. [Apache Otava (Incubating)](https://otava.incubator.apache.org/). Read October 29, 2026, 21:39 UTC.
29. [otava release 0.8.0-incubating on GitHub](https://github.com/apache/otava/releases/tag/0.8.0-incubating). Read October 29, 2026, 21:19 UTC.
30. [Apache OzHera (Incubating)](https://ozhera.incubator.apache.org/). Read October 29, 2026, 21:13 UTC.
31. [Apache SkyWalking](http://skywalking.apache.org). Read October 29, 2026, 21:13 UTC.
32. [Incubator guidance](https://incubator.apache.org/guides/branding.html). Read October 29, 2026, 21:39 UTC.
33. [BLS administrator wage](https://data.bls.gov/oesprofile/?areas=INDUSTRY,STATE,MSA&major_group=150000&measure=01&occupation=151244&year=2025). Read October 29, 2026, 21:12 UTC.
34. [70.0% wages share of compensation](https://data.bls.gov/timeseries/CMU2020000000000P). Read October 29, 2026, 21:17 UTC.
35. [DigitalOcean monthly prices](https://www.digitalocean.com/pricing/droplets). Read October 29, 2026, 21:10 UTC.
36. [Backblaze B2](https://www.backblaze.com/cloud-storage/pricing). Read October 29, 2026, 21:10 UTC.
37. [Datadog Infrastructure Pro](https://www.datadoghq.com/pricing/). Read October 29, 2026, 21:38 UTC.
38. [Datadog APM](https://www.datadoghq.com/pricing/?product=application-performance-monitoring). Read October 29, 2026, 21:39 UTC.
39. [ASF security guidance](https://www.apache.org/security/). Read October 29, 2026, 21:39 UTC.
40. [Flume advisories](https://flume.apache.org/security.html). Read October 29, 2026, 21:39 UTC.
41. [Log4j security notices](https://logging.apache.org/log4j/2.x/security.html). Read October 29, 2026, 21:38 UTC.
42. [HertzBeat community help](https://hertzbeat.apache.org/docs/community/contact/). Read October 29, 2026, 21:39 UTC.
43. [Apache License 2.0](https://www.apache.org/licenses/LICENSE-2.0). Read October 29, 2026, 21:39 UTC.

44. [Claude plans and Claude Code](https://claude.com/pricing). Observed October 29, 2026, 23:31:08 UTC.
45. [ChatGPT plans](https://chatgpt.com/pricing/). Read October 29, 2026, about 23:27 UTC in a regular browser.
46. [Codex with ChatGPT plans](https://help.openai.com/en/articles/11369540-using-codex-with-your-chatgpt-plan). Observed October 29, 2026, 23:31:16 UTC.
47. [Codex CLI](https://developers.openai.com/codex/cli). Observed October 29, 2026, 23:32:01 UTC.
48. [Desktop Commander pricing](https://desktopcommander.app/pricing/). Observed October 29, 2026, 23:31:08 UTC.
49. [Desktop Commander](https://desktopcommander.app/). Observed October 29, 2026, 23:31:09 UTC.
50. [Desktop Commander local server license](https://raw.githubusercontent.com/wonderwhy-er/DesktopCommanderMCP/main/LICENSE). Observed October 29, 2026, 23:31:40 UTC.
51. [Desktop Commander server instructions](https://raw.githubusercontent.com/wonderwhy-er/DesktopCommanderMCP/main/README.md). Observed October 29, 2026, 23:31:41 UTC.
52. [Claude in Chrome](https://support.claude.com/en/articles/12012173-get-started-with-claude-in-chrome). Observed October 29, 2026, 23:31:11 UTC.
53. [Claude in Chrome permissions](https://support.claude.com/en/articles/12902446-claude-in-chrome-permissions-guide). Observed October 29, 2026, 23:31:09 UTC.
54. [ChatGPT cloud browser](https://help.openai.com/en/articles/20001280-using-cloud-browser-in-chatgpt). Observed October 29, 2026, 23:32:01 UTC.
55. [Claude API pricing](https://platform.claude.com/docs/en/about-claude/pricing). Observed October 29, 2026, 23:31:14 UTC.
56. [OpenAI API pricing](https://developers.openai.com/api/docs/pricing). Observed October 29, 2026, 23:31:13 UTC.
57. [OpenAI prompt caching](https://developers.openai.com/api/docs/guides/prompt-caching). Observed October 29, 2026, 23:32:00 UTC.
58. [METR experienced developer study](https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/). Observed October 29, 2026, 23:30:58 UTC.
59. [Anthropic consumer terms](https://www.anthropic.com/legal/consumer-terms). Observed October 29, 2026, 23:31:08 UTC.
60. [OpenAI terms of use](https://openai.com/policies/terms-of-use/). Read October 29, 2026, about 23:27 UTC in a regular browser.

*Apache, the Apache feather logo and the Apache project names are trademarks of The Apache Software Foundation. This site is not affiliated with or endorsed by the ASF. Other product names are used for identification only.*