GitHub Copilot coding agent
Ranked 18 of 41 coding agents by recommenders
People recommend it for running coding tasks in an isolated container and reviewing results in one PR. Main caveat: critics say it changes silently and edits PRs unprompted.
People recommend it for running coding tasks in an isolated container and reviewing results in one PR. Main caveat: critics say it changes silently and edits PRs unprompted.
52 people gave an opinion
- recommend it
- 16
- criticize it
- 36
- approval share
- 31%
Recommended in 8 of 13 months
- November 2025: recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: not recommended
- July 2026: recommended
- August 2026: not recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 16 people recommend itNeeds 10 or more, met.
- 0.4 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 8 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Gets most changes right
“it's kind of wild seeing how diligent it is even with the most trivial of changes”
Comment from 6 Feb 2026 in “Claude Code Agent Teams” “I'd say it gets them right the vast majority of the time now.”
Comment from 6 Jan 2026 in “Opus 4.5 is going to change everything” Less setup than alternatives
“Copilot Agent and Claude Code use their own sandbox, which requires less setup but is also quite limited”
Comment from 5 Jan 2026 in “Claude Code On-the-Go” “Jules and GitHub Copilot Agent suffice for similar workflows with less setup.”
Comment from 5 Jan 2026 in “Claude Code On-the-Go”
What people warn about
Each mark is one comment
Inserts unrelated content into PRs
“No one, anywhere, ever wants this or anything like it. Do not inject anything that is outside of the context of the session, ever.”
Comment from 30 Mar 2026 in “Copilot edited an ad into my PR” “irrelevant modification of the PR description to insert unrelated content is specifically egregious”
Comment from 30 Mar 2026 in “Copilot edited an ad into my PR” “It literally went in and edited the PR description 8 minutes after the user submitted it.”
Comment from 31 Mar 2026 in “GitHub backs down, kills Copilot pull-request ads after backlash” - Comment from 31 Mar 2026 in “Copilot edited an ad into my PR”
- Comment from 1 Apr 2026 in “Copilot edited an ad into my PR”
Changes behavior without announcement
“it seems like almost every other week they silently change something with zero public announcement that creates real disruption for our team”
Comment from 21 Aug 2026 in “The August 17 outage, and the work ahead” “they silently change something with zero public announcement or documentation”
Comment from 28 Sep 2026 in “Coding is not solved” Wrong fixes and hallucinated issues
“The copilot agent then implemented that by adding a replace statement, which is not how you are supposed to do this.”
Comment from 8 Feb 2026 in “GitHub Agentic Workflows” “this actually doesn't work, even though the documentation makes it seem like it should.”
Comment from 9 Feb 2026 in “Billing can be bypassed using a combo of subagents with an agent definition” “it caught GitHub's own copilot-swe-agent hallucinating a fake RCE vulnerability”
Comment from 4 Mar 2026 in “Show HN: The Janitor – A 58MB Rust static analyzer to block AI-generated PR slop” Limited sandbox and slow
“Copilot Agent and Claude Code use their own sandbox, which requires less setup but is also quite limited”
Comment from 5 Jan 2026 in “Claude Code On-the-Go” “I found that to be slow compared to natively running copilot in the IDE.”
Comment from 29 Dec 2025 in “I program on the subway”
Other coding agents people recommend
The four most recommended other coding agents, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| GitHub Copilot | 571 | 993 | 37% | 12 | 1 yr46 owners | — |
| oh-my-pi | 104 | 28 | 79% | 9 | 2 mo5 owners | Approved |
| Grok Build CLI | 66 | 114 | 37% | 6 | — | — |
| DeepSeek Harness | 63 | 28 | 69% | 4 | 1 mo5 owners | Approved |
| GitHub Copilot coding agentThis page | 16 | 36 | 31% | 8 | — | — |
What people said
40 of the 56 comments that mention it, newest first. Each links to the original comment.
“they silently change something with zero public announcement or documentation”
“it seems like almost every other week they silently change something with zero public announcement that creates real disruption for our team”
“copilot's agents tab spared me”
“Microsoft's answer to this was to just ... expose none of this to their Agent Mode!?”
“it's in its own container/VM naturally and you can spawn from the Github Mobile app”
“The copilot cloud agent is great”
“You're impersonating other people to post your advertising in their name”
“It literally went in and edited the PR description 8 minutes after the user submitted it.”
“So you continue to show ads to Copilt, just not to the user? If so, not a fix.”
“it would be generous to say you're stretching the truth here”
“I don't know what other parts of your comment are honest if the core statement is false”
“You guys are disingenuous actors. Bad faith and all that.”
“Never ever trust them because their words are completely empty and they will never change.”
“No one, anywhere, ever wants this or anything like it. Do not inject anything that is outside of the context of the session, ever.”
“I don't think this problem can be solved while leadership is focused only on adding more slop.”
“Whoever did this must have realised the users will hate it.”
“When you do things like this, you lose trust. Work to get it back”
“irrelevant modification of the PR description to insert unrelated content is specifically egregious”
“it felt significantly more robust”
“I've had plenty of success with greenfield projects myself but using the copilot agent”
“Github Copilot Agentic Mode + Sonnet 4.6 is stable and inexpensive.”
“MS GitHub copilot agents are so awful at MS PowerShell”
“it caught GitHub's own copilot-swe-agent hallucinating a fake RCE vulnerability”
“I can (and do) use GH Copilot's "agent" mode with older generation models, and it's fine.”
“this actually doesn't work, even though the documentation makes it seem like it should.”
“The copilot agent then implemented that by adding a replace statement, which is not how you are supposed to do this.”
“it's kind of wild seeing how diligent it is even with the most trivial of changes”
“Neither Copilot Coding Agent nor Claude Code Web can do this.”
“But what I like about this setup is that I have almost all the context I need to review the work in a single PR.”
“It's annoying but I don't think there is a way around having to do that.”
“Go to the URL, type what you want done, and a cloud Claude agent creates a PR.”
“I'd say it gets them right the vast majority of the time now.”
“Copilot Agent and Claude Code use their own sandbox, which requires less setup but is also quite limited”
“Jules and GitHub Copilot Agent suffice for similar workflows with less setup.”
“I found that to be slow compared to natively running copilot in the IDE.”
“so I can evaluate the change works really well”
“the agent mode of copilot will search for and be pretty good and filling its own context afaik”
“I'd replace them with the Github Copilot built-in agent in a heartbeat”
“But it's currently in a very primitive state.”
“The VSC agent mode can do everything you want an agent to do”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.