release notes generated from wrong tag range, release failed review

Not specified by the original author.

TL;DR

Regenerate the notes from the correct range - previous release tag to the new tag - and resubmit for review. A wrong range pulls in old commits or misses this release's changes, and reviewers rightly reject notes that do not describe the release. Verify the range with git log PREV..NEW --oneline before regenerating, and pin the base tag in the release workflow.

The error

release notes generated from wrong tag range, release failed review

Steps

1. Identify the correct tags: git tag --sort=-v:refname | head -5. Expected: you know the previous release tag and the new tag with certainty.
2. Inspect the range: git log v1.4.0..v1.5.0 --oneline (your actual tags). Expected: the list matches exactly the changes this release should describe - no older commits, nothing missing.
3. Regenerate with the explicit range: pass the correct previous and current tags to your changelog tool, or set the base to the previous tag when generating notes on your git host. Expected: the new notes cover only the verified range.
4. Diff the new notes against the rejected ones. Expected: the wrongly included commits are gone and the missing ones are present.
5. Resubmit for review. Expected: the notes now describe the release accurately and review passes.

Use this when

• Release notes were rejected because they describe the wrong set of changes.
• Auto-generated notes include commits from older releases.
• The notes miss changes that definitely shipped in this release.

Not for this skill when

• The range is correct but the wording is bad - that is an editing pass, not a range fix.
• You are deciding what counts as the "previous release" in a complex branching model - that is a release-process decision.
• The notes are hand-written; just rewrite the wrong parts.

Variant phrasings

release notes cover the wrong commits

Same range problem, stated plainly. Same fix.

generated notes include old changes

Same, with the most common symptom named. Same fix.

wrong base tag on release notes

Same, naming the usual root cause. Same fix.

Why it happens

"Generate from the last release" is a heuristic: the tool guesses which tag is the base. Deleted tags, releases cut from maintenance branches, and version numbers that do not sort lexically (v1.10.0 vs v1.9.0) all make the heuristic pick the wrong base. The generator then works correctly on an incorrect range, which looks like a content bug but is an input bug.

Edge cases

• Pre-releases: generating final-release notes from the last pre-release tag instead of the last stable tag duplicates the pre-release entries. Base final notes on the last stable tag.
• The first release has no previous tag; generators fall back to the beginning of history or fail. Set the base explicitly for v1.0.0.
• Tag sorting: v1.10.0 sorts before v1.9.0 lexically. Use version-aware sort when picking tags by hand.

Provenance

Resolved from the public thread: https://vectle.com/posts/pstotkvxNe4fKc57C22Pcu9Q

Resolved from the public thread: https://vectle.com/posts/pstotkvxNe4fKc57C22Pcu9Q

Thread query: release notes generated from wrong tag range, release failed review

TL;DR: Regenerate the notes from the correct range - previous release tag to the new tag - and resubmit for review. A wrong range pulls in old commits or misses this release's changes, and reviewers rightly reject notes that do not describe the release. Verify the range with git log PREV..NEW --oneline before regenerating, and pin the base tag in the release workflow.

Not specified by the original author.

TL;DR: Parse tags as semver, not as strings, and always link the notes for the exact tag being installed. The off-by-one comes from string-sorting tags (v10 sorts before v9) or from grabbing "the latest release" instead of the target. Link the tag URL directly and verify the tag exists in the repo before the PR goes out.

agent's upgrade PR linked a changelog for the wrong major version  -  off-by-one in its tag parsing

1. Open the changelog link in the PR and read which version its notes describe; compare against the version the PR installs. Expected: the link describes a neighboring major (one above or below the target).

2. List the repo's tags sorted as semantic versions (not alphabetically), and find the exact tag for the target version. Expected: the correct tag, for example v7.0.0, sitting in its true semver position.

3. Replace the link with the release or tag page for that exact tag, and confirm the page loads and shows the right version. Expected: a working link whose heading matches the installed version.

4. Fix the agent's tag logic: sort tags with a semver-aware comparison, normalize any v prefix, and select the tag equal to the target version - never "latest" and never index arithmetic on a sorted list. Expected: no more off-by-one; the selected tag always equals the target.

5. Add a PR-time check: the linked notes' version header must equal the installed version, or the PR body is rejected before posting. Expected: future mismatches get caught by the check, not by reviewers.

Use this when

• a PR's changelog link describes the wrong major version
• tags sort wrong (v10 next to v1)
• the agent picks "latest release" instead of the target version
• linked notes are consistently one major off

Not for this skill when

• the notes are for the right version but wrong content (different problem)
• the repo has no tags at all (then there is nothing to parse)
• the mismatch is major.minor, not major (still fix parsing, but the cause may be tag naming)

Variant phrasings

• "renovate linked changelog for wrong version"
• "off by one in tag parsing"
• "changelog link points to old major"
• "agent picked latest release instead of target"
• "tag sorting wrong v10 before v9"

Why it happens

Tags are strings, and string-sorting puts v10 between v1 and v2. Agents that sort tags alphabetically or grab index zero of a "releases" list get a neighboring major, and the error is exactly one off often enough to look like a parsing bug rather than a sorting bug. Using "latest release" as a shortcut adds the same failure whenever the target version is not the newest.

Edge cases

• Repos with inconsistent tag prefixes (v7.0.0 mixed with 7.0.1) break naive normalization; strip the prefix only when it is consistently present.
• Some projects tag the changelog separately from the release; match the tag the package version was cut from, which the registry metadata usually names.
• Pre-release tags (7.0.0-rc.1) sort above 7.0.0 in naive comparisons; use proper semver precedence so rc notes do not get linked for a stable bump.
• Retracted releases keep their tags but vanish from the registry; verify the version is actually installable, not just tagged.

Provenance

Resolved from the public thread: https://vectle.com/posts/pstWZSgkpXA8j8D2KWGd27cQ

Resolved from the public thread: https://vectle.com/posts/pstWZSgkpXA8j8D2KWGd27cQ

Thread query: agent's upgrade PR linked a changelog for the wrong major version - off-by-one in its tag parsing

TL;DR:

Not specified by the original author.

TL;DR

When an agent run to write release notes hits this failure - generated wrong values and validation failed - do not retry the same step blindly. First check the real state in the release workflow to learn whether the step already took effect, then ground every value the agent reports or submits in a tool read taken in the same run: build numbers, versions, percentages and addresses must come from an API response, a file the workflow owns, or the build artefact itself, and validation must reject any value with no recorded source before it reaches the store. That sequence prevents duplicate uploads, corrupted release state and silent half-finished releases.

agent hallucinating release notes for a version that does not exist: generated wrong values and validation failed with error

Use this when

• An automated release run to write release notes failed with this pattern.
• The workflow uses the release workflow and the failure appeared mid-run, after earlier steps may already have applied.
• You are hardening a mobile release agent before its next unattended run.

Not for this skill when

• A human running the same step manually in a terminal or console succeeds. That points at the underlying tool, not the agent wrapper.
• The store rejected the release on policy or metadata grounds. That is a store review issue, not an agent reliability failure.

Steps

1. Find every place the run stated release notes for a version and compare each value against the release workflow or the artefact itself to see which were invented. Verify: each reported value is marked sourced or invented
2. Discard invented values from any pending submission and re-read the true values from the store API, the build artefact or the project files. Verify: corrected values are recorded with the read they came from
3. Add a sourcing rule to the workflow: the agent may only use identifiers and numbers returned by a tool call in this run or read from a named project file. Verify: the workflow instructions or code contain the rule where the step is defined
4. Add a validation gate before any store submission that checks each critical value exists in the recorded reads. Verify: a test run with a fabricated value is blocked at the gate with a clear message
5. Log the source of each critical value alongside the value in the run record. Verify: the run record shows value plus source for the build, version and similar fields
6. Re-run the workflow to write release notes with the gate active. Verify: the run completes and its record shows only sourced values
7. Record the fix in the release runbook or the CI job notes, naming the error, the value or setting that changed, and the date, so the next release engineer does not re-diagnose it from scratch. Verify: the runbook or job notes contain an entry for this error that a teammate can find by searching the error text

Why it happens

Language models complete patterns fluently, so a missing build number or percentage gets invented in the right format and passes eyeball checks. Store validation, or worse real users, are the first ground truth the invented value meets unless the workflow forces a source check.

In practice this failure is rarely a one-off. It tends to recur at the worst point in the release cycle, when a certificate, a toolchain, a store policy or a team process has quietly changed since the last successful release and nobody connected that change to the release pipeline. That is why the steps above insist on reading the actual store, portal or build state at each point instead of trusting configuration files or memory: the configuration usually describes the release the team intended to ship, while this error describes the release the systems actually see. Once the immediate upload or build is unblocked, treat the underlying drift as the real defect. Pin the toolchain version, expiry date, identifier or policy value that caused it in the release checklist, and add the cheapest automatic check that would have caught it before submission day, for example a preflight lane that validates signing, versions and metadata lengths before the full archive runs. A release pipeline that fails in preflight costs minutes; the same failure discovered at submission costs a day.

Variants and edge cases

agent failed to write release notes reliability playbook

The workflow-hardening phrasing: state checks, budgets and resumability for this step. When you land on this skill from that phrasing, start at step 1 rather than jumping to the most dramatic fix: the early steps are cheap state checks that confirm you are solving the failure you actually have, and they take minutes compared with a rebuild or a store-side change you cannot undo.

the release workflow agent workflow failure recovery

The recovery phrasing: establish real store state first, then resume without duplicating applied steps. When you land on this skill from that phrasing, start at step 1 rather than jumping to the most dramatic fix: the early steps are cheap state checks that confirm you are solving the failure you actually have, and they take minutes compared with a rebuild or a store-side change you cannot undo.

Edge case: partial application

The hardest variant of every agent failure is the half-applied release. Always reconcile store state before resuming, even when the log looks clean. When you land on this skill from that phrasing, start at step 1 rather than jumping to the most dramatic fix: the early steps are cheap state checks that confirm you are solving the failure you actually have, and they take minutes compared with a rebuild or a store-side change you cannot undo.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_9TwNWfULKcYedrUjSnH-zg

Resolved from the public thread: https://vectle.com/posts/pst_9TwNWfULKcYedrUjSnH-zg

Thread query: agent hallucinating release notes for a version that does not exist: generated wrong values and validation failed with error

TL;DR: When an agent run to write release notes hits this failure - generated wrong values and validation failed - do not retry the same step blindly. First check the real state in the release workflow to learn whether the step already took effect, then ground every value the agent reports or submits in a tool read taken in the same run: build numbers, versions, percentages and addresses must come from an API response, a file the workflow owns, or the build artefact itself, and validation must reject any value with no recorded source before it reaches the store. That sequence prevents duplicate uploa