issue

Manage Sentry issues

Commands

sentry issue list [<org/project>]

List issues in a project

Arguments:

Argument	Description
[<org/project>]	<org>/ (all projects), <org>/<project>, or <project> (search)

Options:

Option	Description
-q, --query <query>	Search query (Sentry syntax, implicit AND; same-key OR becomes an in-list)
-n, --limit <limit>	Maximum number of issues to list (default: "25")
-s, --sort <sort>	Sort by: recommended, date, new, freq, user (default: recommended on sentry.io, else date)
-t, --period <period>	Time range: "7d", "2024-01-01..2024-02-01", ">=2024-01-01" (default: "90d")
-c, --cursor <cursor>	Pagination cursor (use "next" for next page, "prev" for previous)
--compact	Single-line rows for compact output (auto-detects if omitted)
-f, --fresh	Bypass cache, re-detect projects, and fetch fresh data

Examples:

# List issues (project auto-detected)
sentry issue list

# List issues in a specific project
sentry issue list my-org/frontend

# Several projects in the same organization
sentry issue list my-org/frontend,backend,worker

# All projects in an organization
sentry issue list my-org/

# Search for a project across organizations
sentry issue list frontend

# Show only unresolved issues
sentry issue list my-org/frontend --query "is:unresolved"

# Sort by frequency
sentry issue list my-org/frontend --sort freq --limit 20

# Sort by Sentry's recommended relevance ranking
sentry issue list my-org/frontend --sort recommended

# Multiple filters (space-separated terms are combined with AND)
sentry issue list --query "is:unresolved level:error assigned:me"

# Negation and wildcards
sentry issue list --query "!browser:Chrome message:*timeout*"

# Match multiple values for one key
sentry issue list --query "browser:[Chrome,Firefox]"

sentry issue events <issue>

List events for a specific issue

Arguments:

Argument	Description
<issue>	Issue: @latest, @most_frequent, <org>/ID, <org>/<project>#ID, <project>-suffix, ID, or suffix

Options:

Option	Description
-n, --limit <limit>	Number of events (1-1000) (default: "25")
-q, --query <query>	Search query (Sentry search syntax)
--full	Include full event body (stacktraces)
-t, --period <period>	Time range: "7d", "2024-01-01..2024-02-01", ">=2024-01-01" (default: "7d")
-f, --fresh	Bypass cache, re-detect projects, and fetch fresh data
-c, --cursor <cursor>	Navigate pages: "next", "prev", "first" (or raw cursor string)

Examples:

# List recent events for an issue
sentry issue events FRONT-ABC

# List events for the most recent issue
sentry issue events @latest --limit 50

# Filter events by search query
sentry issue events FRONT-ABC --query "browser:Chrome"

# Show full event details
sentry issue events FRONT-ABC --full

# Limit results and filter by time period
sentry issue events FRONT-ABC --limit 50 --period 24h

# Fetch the next page
sentry issue events FRONT-ABC -c next

# Output as JSON
sentry issue events FRONT-ABC --json

sentry issue tags <issue> <key>

Show the top values of an issue tag

Arguments:

Argument	Description
<issue>	Issue: @latest, @most_frequent, <org>/ID, <org>/<project>#ID, <project>-suffix, ID, or suffix
<key>	Tag key, e.g. release, server_name, browser.name

Options:

Option	Description
-e, --environment <environment>	Filter to one environment (default: all environments)
-f, --fresh	Bypass cache, re-detect projects, and fetch fresh data

Examples:

# Show which releases concentrate an issue's events
sentry issue tags my-org/FRONT-ABC release

# Inspect server distribution for the latest issue in production
sentry issue tags @latest server_name --environment production

# Get browser counts and timestamps as JSON
sentry issue tags FRONT-ABC browser.name --json

# Select the aggregate fields for a release distribution
sentry issue tags FRONT-ABC release --json --fields key,totalValues,topValues

sentry issue explain <issue...>

Analyze one or more issues using Seer AI

Arguments:

Argument	Description
<issue...>	One or more issue IDs

Options:

Option	Description
--force	Force new analysis even if one exists
-f, --fresh	Bypass cache, re-detect projects, and fetch fresh data

Examples:

# Analyze the root cause of an issue (may take a few minutes for new issues)
sentry issue explain 123456789

# Explain the most recent issue
sentry issue explain @latest

# Explain the most frequently occurring issue
sentry issue explain @most_frequent

# Use a short ID with an organization prefix
sentry issue explain my-org/MYPROJECT-ABC

# Analyze multiple issues in one invocation
sentry issue explain FRONT-ABC BACK-2

# Force a fresh analysis
sentry issue explain 123456789 --force

# Output root causes as JSON
sentry issue explain 123456789 --json

sentry issue plan <issue>

Generate a solution plan using Seer AI

Arguments:

Argument	Description
<issue>	Issue: @latest, @most_frequent, <org>/ID, <org>/<project>#ID, <project>-suffix, ID, or suffix

Options:

Option	Description
--force	Force new plan even if one exists
-f, --fresh	Bypass cache, re-detect projects, and fetch fresh data

Examples:

# Generate a fix plan (runs explain first if needed)
sentry issue plan 123456789

# Generate a fix plan for the most recent issue
sentry issue plan @latest

# Use a short ID with an organization prefix
sentry issue plan my-org/MYPROJECT-ABC

# Force a fresh plan even if one already exists
sentry issue plan 123456789 --force

sentry issue view <issue...>

View details of one or more issues

Arguments:

Argument	Description
<issue...>	One or more issue IDs

Options:

Option	Description
-w, --web	Open in browser
--force	Allow --web to open more than 5 issues
--spans <spans>	Span tree depth limit (number, "all" for unlimited, "no" to disable) (default: "3")
-f, --fresh	Bypass cache, re-detect projects, and fetch fresh data

Examples:

# View an issue
sentry issue view FRONT-ABC

# View the most recent issue
sentry issue view @latest

# View several issues (space-separated, not commas)
sentry issue view FRONT-ABC BACK-2

# Select an organization explicitly
sentry issue view my-org/FRONT-ABC

# Use a GitHub-style identifier (the "#" replaces the final slash)
sentry issue view my-org/my-project#FRONT-ABC

# Open issues in the browser (up to 5 tabs by default)
sentry issue view FRONT-ABC BACK-2 -w

# Allow more than 5 browser tabs
sentry issue view FRONT-ABC BACK-2 API-3 WEB-4 IOS-5 OPS-6 -w --force

# Full JSON (issue fields + latest event + trace/replay context)
sentry issue view FRONT-ABC --json

# Select top-level fields to keep output small
sentry issue view FRONT-ABC --json --fields shortId,title,culprit,count,userCount,permalink

# Select fields from the latest event instead of the whole event
sentry issue view FRONT-ABC --json --fields event.id,event.title,event.dateCreated

# Extract the request URL and method from the latest event
sentry issue view FRONT-ABC --json | jq '.event.entries[] | select(.type == "request") | .data | {url, method}'

# Extract the exception type and value from the latest event
sentry issue view FRONT-ABC --json | jq '.event.entries[] | select(.type == "exception") | .data.values[0] | {type, value}'

sentry issue resolve <issue>

Mark an issue as resolved

Arguments:

Argument	Description
<issue>	Issue: @latest, @most_frequent, <org>/ID, <org>/<project>#ID, <project>-suffix, ID, or suffix

Options:

Option	Description
-i, --in <in>	Resolve in a release, next release, or commit (''

Examples:

# Resolve immediately (no regression tracking)
sentry issue resolve CLI-G5

# Resolve in a specific release (newer releases flag regressions)
sentry issue resolve CLI-G5 --in 0.26.1

# Monorepo-style release versions work too
sentry issue resolve CLI-G5 --in spotlight@1.2.3

# Resolve in the next release
sentry issue resolve CLI-G5 --in @next

# Resolve in the current git HEAD (auto-detects the Sentry repo from your origin remote)
sentry issue resolve CLI-G5 --in @commit

# Resolve in an explicit commit and repo (no git inspection)
sentry issue resolve CLI-G5 --in @commit:getsentry/cli@abc123def

# Resolve an issue in an explicit organization
sentry issue resolve my-org/CLI-G5

sentry issue unresolve <issue>

Reopen a resolved issue

Arguments:

Argument	Description
<issue>	Issue: @latest, @most_frequent, <org>/ID, <org>/<project>#ID, <project>-suffix, ID, or suffix

Examples:

# Reopen a resolved issue
sentry issue unresolve CLI-G5

# `reopen` is an alias for `unresolve`
sentry issue reopen CLI-G5

# Reopen an issue in an explicit organization
sentry issue unresolve my-org/CLI-G5

sentry issue archive <issue>

Archive (ignore) an issue

Arguments:

Argument	Description
<issue>	Issue: @latest, @most_frequent, <org>/ID, <org>/<project>#ID, <project>-suffix, ID, or suffix

Options:

Option	Description
-u, --until <until>	Condition for unarchival: forever, auto, 30m, 10x, 10u, 10x/5m, etc.

Examples:

# Archive forever (fully silenced)
sentry issue archive CLI-G5

# Unarchive when Sentry detects a spike in event frequency
sentry issue archive CLI-G5 --until auto

# Archive for 1 hour
sentry issue archive CLI-G5 --until 1h

# Archive for 7 days
sentry issue archive CLI-G5 --until 7d

# Archive until a specific date
sentry issue archive CLI-G5 --until 2026-12-31

# Unarchive after 100 more events
sentry issue archive CLI-G5 --until 100x

# Unarchive after 10 more users are affected
sentry issue archive CLI-G5 --until 10u

# Unarchive after 100 events within 1 hour
sentry issue archive CLI-G5 --until 100x/1h

# Unarchive after 10 users within 1 day
sentry issue archive CLI-G5 --until 10u/1d

# Use the verbose condition form
sentry issue archive CLI-G5 --until 10events/2hours

# `ignore` is an alias for `archive`
sentry issue ignore CLI-G5 --until auto

sentry issue merge <issue...>

Merge 2+ issues into a single canonical group

Destructive

This command deletes data or makes changes that can't easily be undone.

Arguments:

Argument	Description
<issue...>	Issue IDs to merge (2 or more required)

Options:

Option	Description
-i, --into <into>	Prefer this issue as the canonical parent (included in the merge if not already listed)

Examples:

# Let Sentry pick the parent (typically the largest by event count)
sentry issue merge CLI-K9 CLI-15H CLI-15N

# Pin the canonical parent explicitly
sentry issue merge CLI-K9 CLI-15H CLI-15N --into CLI-K9

# Use organization-qualified IDs
sentry issue merge my-org/CLI-K9 my-org/CLI-15H --into my-org/CLI-K9

# Use project-alias IDs
sentry issue merge cli-k9 cli-15h --into cli-k9

sentry issue link <issue> <url>

Link an existing external issue

Arguments:

Argument	Description
<issue>	Issue: @latest, @most_frequent, <org>/ID, <org>/<project>#ID, <project>-suffix, ID, or suffix
<url>	URL of an existing tracker issue or GitHub pull request

Options:

Option	Description
--integration <integration>	Native integration ID, when multiple installations match
--app <app>	Sentry App slug (automatically detected for Linear URLs)
-n, --dry-run	Show what would happen without making changes
--field <field>...	Additional Sentry App link form field (name=value, repeatable)

Examples:

# Link a GitHub issue
sentry issue link FRONT-123 https://github.com/example/app/issues/42

# Link a GitHub pull request
sentry issue link FRONT-123 https://github.com/example/app/pull/43

# Link a Jira issue in an explicit organization
sentry issue link my-org/FRONT-123 https://example.atlassian.net/browse/APP-42

# Link a Linear issue
sentry issue link FRONT-123 https://linear.app/example/issue/APP-42/fix-error

# Prepare the link without submitting it
sentry issue link my-org/FRONT-123 https://github.com/example/app/issues/42 --dry-run

# Output the result as JSON
sentry issue link my-org/FRONT-123 https://github.com/example/app/issues/42 --json

sentry issue unlink <issue> <url>

Unlink an external issue

Destructive

This command deletes data or makes changes that can't easily be undone. Use --dry-run to preview the changes first.

Arguments:

Argument	Description
<issue>	Issue: @latest, @most_frequent, <org>/ID, <org>/<project>#ID, <project>-suffix, ID, or suffix
<url>	URL of an existing tracker issue or GitHub pull request

Options:

Option	Description
--integration <integration>	Native integration ID, when multiple installations match
--app <app>	Sentry App slug (automatically detected for Linear URLs)
-y, --yes	Skip confirmation prompt
-f, --force	Force the operation without confirmation
-n, --dry-run	Show what would happen without making changes

Examples:

# Unlink a GitHub issue (prompts for confirmation)
sentry issue unlink FRONT-123 https://github.com/example/app/issues/42

# Unlink a GitHub pull request without prompting
sentry issue unlink FRONT-123 https://github.com/example/app/pull/43 --yes

# Unlink a Jira issue in an explicit organization
sentry issue unlink my-org/FRONT-123 https://example.atlassian.net/browse/APP-42 --yes

# Check whether a link exists without removing it
sentry issue unlink FRONT-123 https://linear.app/example/issue/APP-42/fix-error --dry-run

All commands support --json for machine-readable output and --fields to select specific JSON fields.

Context recovery

If auto-detection cannot find a project and your account can reach exactly one organization, issue list lists that organization's issues, the same as sentry issue list <org>/.

With several organizations, authenticated non-interactive runs show one copyable retry example and compact lists of other available organizations and org/project targets. Choose the project that matches your task; the retry keeps your query, period, limit, sort, and output flags. Navigation keywords such as --cursor next become --cursor first because the chosen project may have no saved pagination state; raw API cursors are preserved. Suggestions are a bounded sample, with one sentry project list example to discover other projects. Showing suggestions does not save defaults.

Search syntax

Search syntax

Sentry search uses implicit AND — space-separated terms are all required. AND/OR operators are not supported for issue search. Use alternatives:

• key:[val1,val2] — in-list syntax (matches val1 OR val2 for one key)
• status:[resolved,ignored] or substatus:[regressed,escalating] — is: has no in-list form
• Run separate queries for different terms
• *term* — wildcard matching

The CLI rewrites same-key ORs to in-lists automatically, including is:resolved OR is:ignored → status:[resolved,ignored].

Full syntax reference: Sentry Search Docs

Magic selectors

Use @latest and @most_frequent to target issues without knowing their ID.

Common mistakes

• Don't use the issue command name as an issue prefix — sentry issue explain issue-1 is parsed as project issue + suffix 1. Use a real project prefix: sentry issue explain FRONT-1 or a numeric ID: sentry issue explain 123456789.
• Issue short IDs use uppercase prefixes — CLI-G, not cli-g. The CLI tolerates lowercase input in most commands, but agents should prefer the canonical form.
• org/project targets use slugs, not numeric IDs — sentry issue list my-org/6775615880 fails when 6775615880 is a project ID copied from a URL. Run sentry project list my-org/ to find the slug (e.g. frontend).
• Prefer auto-detect over guessing slugs — omit the target when possible; the CLI resolves org/project from DSNs, .env files, and your working directory.

Inspect tag distributions

After opening an issue, compare which releases, servers, or browsers account for its events. issue view shows tags on the latest event; issue tags aggregates across the issue's retained events.

The result contains up to 9 top values, ordered by count. Human output shows each value's count and percentage of all retained events, including events with a missing value. Any remainder appears as "Other values (not shown)"; percentages are never calculated from just the displayed rows. Retention limits apply, so this total can differ from the issue's lifetime event count.

--environment selects one environment; omitting it includes all environments. This aggregate does not support --period, --query, --limit, or cursor pagination. Use issue events when you need filtered event records. For example, after identifying a release, inspect recent events with sentry issue events FRONT-ABC --query 'release:1.2.3' --period 7d.

JSON preserves key, name, totalValues, and topValues, including counts and first/last seen timestamps. Values and counts may be null; missing data is not zero. --fields selects top-level fields. To discover tag keys, inspect the latest event with sentry issue view <issue>; it may not contain every tag used by earlier events. A tag with no retained values, or an unknown environment, can return a not-found error.

Short IDs and organizations

Full short IDs such as FRONT-ABC use your configured organization: SENTRY_ORG, then .sentryclirc, then the default set with sentry cli defaults org. The lookup stays within that organization. To select another organization, include it explicitly, as in sentry issue view my-org/FRONT-ABC.

Project aliases such as f-abc keep the organization associated with the alias. Without a configured organization, the CLI uses a matching detected project or searches your accessible organizations. If the short ID matches in multiple organizations, specify the organization explicitly.

JSON output

--json returns the issue fields at the top level plus the latest event under event, the resolved org slug, related replayIds, and trace context. Prefer this over the human output when parsing programmatically. One issue ID still returns a single object; multiple IDs return an array of those objects.

In the --json output, the latest event's data lives under event.entries[], each tagged with a type ("exception", "request", "breadcrumbs", …) and a data payload — not as top-level event.request / event.exception keys.

With --json, one explained issue preserves the existing array of root causes. Multiple issues return an array of labeled objects containing issue, org, issueId, and rootCauses.

Seer AI plans

sentry issue plan generates a fix plan and runs root cause analysis first if needed.

Requirements:

• Seer AI enabled for your organization
• GitHub integration configured with repository access
• Code mappings set up to link stack frames to source files
• Root cause analysis is run automatically if needed (the plan command triggers explain first)

Resolving in a release

How @commit auto-detects

--in @commit reads HEAD and the origin remote, parses the remote as owner/repo, then looks it up in your org's Sentry repositories (cached locally for 7 days). If any step fails, the command stops with a clear error pointing you at --in @commit:<repo>@<sha> or sentry repo list <org>/ — no silent fallback to a different resolution mode.

Merging issues

Consolidate multiple issues (e.g. same logical error split by Sentry's default stack-trace grouping) into a single canonical group with sentry issue merge. Without --into, Sentry picks the parent (typically the largest by event count).

All merged issues must share an organization; cross-org merges are rejected. Non-error issue types (performance, info, etc.) cannot be merged.

Archiving issues

Archive an issue to suppress alerts. Without --until, the issue is archived forever. Use --until to set a condition for automatic unarchival.

--until syntax reference

Format	Meaning
auto	Unarchive on event frequency spike (recommended)
30m, 1h, 7d, 1w	Duration (minutes, hours, days, weeks)
2026-05-15	Absolute date (computed as time delta)
10x or 10events	After 10 more events
10u or 10users	After 10 more users affected
10x/5m	10 events within 5 minutes
10users/2hours	10 users within 2 hours
(omitted)	Archive forever
:::

Linking external issues

sentry issue link links an existing tracker issue or GitHub pull request to a Sentry issue.

The matching integration must already be installed in the Sentry organization. Linking requires a Sentry version with native issue URL resolution and guarded Sentry App callbacks; older self-hosted versions may require an upgrade. Native integrations include GitHub, GitHub Enterprise, Jira, Jira Server, GitLab, Bitbucket, and Azure DevOps. Linear uses its installed Sentry App. Sentry resolves native issue URLs through the selected integration; the remote issue must be visible to that installation. Use --integration <id> if more than one native integration matches the URL. Other Sentry Apps require --app <slug> and must expose an issue-link form; additional required form values can be supplied with --field name=value. For other Apps, an issue select can be supplied by exact ID or label with --field, for example --app custom --field task_id=123. Sentry checks that the app's callback identifies the requested URL before saving the association.

--dry-run discovers the integration and prepares the link without submitting a write. The provider validates the remote issue when the link is submitted. An existing matching link succeeds with changed: false. A Sentry App that already links this issue to a different resource must be unlinked first. App callbacks must return the exact supplied URL; use the issue URL copied from the tracker, including its title suffix. A mismatch fails without saving the link.

GitHub and GitHub Enterprise pull requests are stored as external references. Their /pull/NUMBER and /issues/NUMBER URLs identify the same resource for duplicate detection and unlinking. Linking a PR does not mark it as a fix or resolve the Sentry issue.

This command does not create a tracker issue or link a commit. Existing integration status-sync settings continue to apply after linking.

Link permissions

Linking requires event:write and access to the Sentry project. Discovering Sentry Apps also requires org:read. Both scopes are included in the default OAuth login. If an older OAuth session lacks the requested scopes, the CLI offers reauthorization after a permission error. Use sentry auth login to request the current default scopes. Environment tokens must be updated separately.

Unlinking external issues

sentry issue unlink removes an association without deleting either issue.

Use --yes for non-interactive execution. --dry-run shows whether the link exists without removing it. If the association is already absent, the command succeeds with changed: false.

Unlink matches the URL against stored associations and sends Sentry's internal link ID to the existing DELETE endpoint. It does not require fetching the ticket from the remote tracker, so a deleted remote ticket can still be unlinked. For a custom Sentry App, select it with --app <slug>; unlink does not require the app to expose a link form. Use --integration <id> to disambiguate native integration links.

Unlink permissions

Unlink requires event:write and access to the Sentry project; event:admin is also accepted. The organization's “Let Members Delete Events” setting does not restrict unlinking on updated Sentry versions.

Granting a token more scopes does not override project-access policy.